Catalytic Hydrodeoxygenation of Bio-Oil Model Compounds (Ethanol, 2-

Methyltetrahydrofuran) over Supported Transition Metal Phosphides

Phuong Phuc Nam Bui

Dissertation submitted to the faculty of Virginia Polytechnic Institute and State University in

partial fulfillment of the requirements for the degree of

Doctor of Philosophy

In

Chemical Engineering

David F. Cox, Chair

Shigeo T. Oyama

Y. A. Liu

John Y. Walz

December 20, 2012

Blacksburg, VA

Keywords: transition metal phosphide, nickel phosphide, tungsten phosphide,

hydrodeoxygenation (HDO), deoxygenation, ethanol, 2-methyltetrahydrofuran, FTIR

spectroscopy.

Copyright © 2012, Phuong Phuc Nam Bui

Catalytic Hydrodeoxygenation of Bio-Oil Model Compounds (Ethanol, 2-

Methyltetrahydrofuran) over Supported Transition Metal Phosphides

Phuong Phuc Nam Bui

Abstract

The objective of this project is to investigate hydrodeoxygenation (HDO), a crucial step

in the treatment of bio-oil, on transition metal phosphide catalysts. The study focuses on

reactions of simple oxygenated compounds present in bio-oil – ethanol and 2-

methyltetrahydrofuran (2-MTHF) O . The findings from this project provide fundamental

knowledge towards the hydrodeoxygenation of more complex bio-oil compounds. Ultimately,

the knowledge contributes to the design of optimum catalysts for upgrading bio-oil.

A series of transition metal phosphides was prepared and tested; however, the focus was

on Ni2P/SiO2. Characterization techniques such as X-ray diffraction (XRD), temperature-

programmed reduction and desorption (TPR and TPD), X-ray photoelectron spectroscopy (XPS),

and chemisorption were used. In situ Fourier transform infrared (FTIR) spectroscopy was

employed to monitor the surface of Ni2P during various experiments such as: CO and pyridine

adsorption and transient state of ethanol and 2-MTHF reactions. The use of these techniques

allowed for a better understanding of the role of the catalyst during deoxygenation.

iii

Acknowledgement

The last five years have been an important chapter of my life with some painful loses, but

also many joyful rewards. I am grateful to have taken this journey and met wonderful people

along the way.

I am sincerely grateful to my advisor and mentor Dr. S. Ted Oyama for his guidance and

continuous support throughout the journey. Besides valuable professional critiques and advices,

he taught me the beauty of science (chemistry) and instilled in me a passion for research and

perfection. I also appreciate the time and support provided for my project by my committee

members, Dr. Cox, Dr. Lu, Dr. Liu, and Dr. Walz.

I have gained invaluable experience from the international collaborations during my

program and for this I would like to thank Dr. Juan A. Cecilia, Dr. Antonia Infantes-Molina, Dr.

Enrique Rodríguez-Castellón, Dr. Antonio J. López, Dr. Atsuhi Takagaki, Dr. Ara Cho, Ayako

Iino, Dr. Tao Dou, and Dr. Zhao Shen.

I have special thanks for the invaluable friendships I have been lucky to have with the

group members and other colleagues: Dr. Pelin Hacarlioglu, Dr. Sheima Khatib, Dr. Yujung

Dong, Dr. Dan Li, Dr. Haiyan Zhao, Dr. Chen Chen, and Stuart Higgins. I appreciate the initial

training from Dr. Travis Gott and Dr. Jason Gaudet. I would like to thank the Chemical

Engineering Department staff, Riley Chan, Diane Cannaday, Tina Russell, Michael Vaught,

Kevin Holshouser, Nora Bentley, and Jane Price for their help with the equipment maintenance,

purchase orders, and various paperwork processes.

Words cannot express my gratitude to my large family – my parents, grandparents,

uncles, aunts, and cousins. Though being far away, they have never stopped caring and

iv

encouraging me every step of the way. Special thanks to my Mom and Dad who have been

sacrificing their whole life for a better future for us and to my brother who has been my

motivation.

My dearest love and appreciation goes to my husband – Logan Sorenson, he brings

sunshine to my life, even when the sky is grey. I would also like to extend my thanks to my kind

and sweet in-laws who are my second family.

v

To my parents, Huong Hua and Cuong Bui,

and my life partner, Logan Sorenson,

for their constant love and support

vi

ATTRIBUTION

My dissertation includes collaborations from several colleagues around the world. A brief

description of their contribution is included here.

Chapter 2: Rake Mechanism for the Deoxygenation of Ethanol over a Supported Ni2P/SiO2

Catalyst.

Chapter 2 was published in the Journal of Catalysis.

Dan Li earned his Ph. D from the Key Laboratory of Heavy Oil Processing in China University

of Petroleum. Dr. Li was a co-author on this paper, wrote, and contributed to the research on

catalytic ethanol decomposition.

Haiyan Zhao earned her Ph. D from the Environmental Catalysis and Materials Laboratory,

Virginia Tech. Dr. Zhao was a co-author on this paper, helped with the Weisz-Prater calculation,

and contributed editorial comments.

Shigeo T. Oyama, Ph. D, is a professor in the field of catalysis at both University of Tokyo and

Virginia Tech. Dr. Oyama was a co-author on this paper, principal investigator for the grants

supporting the research, and contributed editorial comments.

Tao Dou, Ph. D, and Zhen Zhao, Ph. D, are professors at the Key Laboratory of Heavy Oil

Processing, China University of Petroleum. Dr. Dou and Dr. Zhao were co-authors on this paper

and contributed editorial comments.

Chapter 3: Synthesis of Transition Metal Phosphides via Phosphate and Phosphite Method and

their Activity in the Hydrodeoxygenation of 2-Methyltetrahydrofuran.

Chapter 3 was published in the Journal of Catalysis.

Juan A. Cecilia, Ph. D, is from the Departamento de Química Inorgánica, Unidad asociada al

ICP-CSIC, Facultad de Ciencias, Universidad de Málaga (Spain). Dr. Cecilia was a co-author on

this paper, wrote, and helped with the catalyst synthesis via phosphite method and the X-ray

photoelectron spectroscopic experiments.

Shigeo T. Oyama, Ph. D, is a professor in the field of catalysis at both the University of Tokyo

(Japan) and Virginia Tech. Dr. Oyama was a co-author on this paper, principal investigator for

one of the grants supporting the research, and contributed editorial comments.

vii

Atsushi Takagaki, Ph. D, is an assistant professor at the University of Tokyo. Dr. Takagaki was a

co-author on this paper and contributed to the X-ray photoelectron spectroscopic experiments.

Antonia Infantes-Molina, Ph. D, is a researcher at the Instituto de Catálisis, CSIC (Spain). Dr.

Infantes-Molina was a co-author on this paper and assisted with the X-ray photoelectron

spectroscopic experiments and editorial comments.

Haiyan Zhao earned her Ph. D from the Environmental Catalysis and Materials Laboratory,

Virginia Tech. Dr. Zhao was a co-author on this paper and contributed editorial comments.

Dan Li earned his Ph. D from the Key Laboratory of Heavy Oil Processing in China University

of Petroleum. Dr. Li was a co-author on this paper and contributed editorial comments.

Enrique Rodríguez-Castellón, Ph. D, and Antonio J. López, Ph. D, are professors at the

Departamento de Química Inorgánica, Unidad asociada al ICP-CSIC, Facultad de Ciencias,

Universidad de Málaga (Spain). Dr. Rodríguez-Castellón and Dr. López helped with the X-ray

photoelectron spectroscopy research and contributed editorial comments.

Chapter 4: Kinetic and Spectroscopic Study of Deoxygenation of 2-Methyltetrahydrofuran over

Supported Nickel Phosphide Catalyst.

Ayako Iino is a graduate student in the Department of Chemical System Engineering of the

University of Tokyo. Iino helped with the temperature-programmed desorption of 2-MTHF from

nickel phosphide.

viii

Contents

Introduction .................................................................................................................... 1 Chapter 1

1.1. Importance of Hydrodeoxygenation..................................................................................... 1

1.2. Catalysts in Hydrodeoxygenation ........................................................................................ 2

1.2.1. Transition metal phosphides .......................................................................................... 3

1.2.2. Preparations of transition metal phosphides .................................................................. 4

1.3. Hydrodeoxygenation of Cyclic Ether Compounds .............................................................. 5

1.3.1. Aromatic cyclic ethers ................................................................................................... 6

1.3.2. Saturated cyclic ethers ................................................................................................. 12

1.4. Conclusions ........................................................................................................................ 16

1.5. Goals................................................................................................................................... 16

1.6. Dissertation Overview ........................................................................................................ 17

Reference ................................................................................................................................... 19

Rake mechanism for the deoxygenation of ethanol over a supported Ni2P/SiO2 catalystChapter 2

....................................................................................................................................................... 23

2.1. Introduction ........................................................................................................................ 23

2.2. Materials and Experiments ................................................................................................. 25

2.2.1. Materials ...................................................................................................................... 26

2.2.2. Nickel phosphide (Ni2P/SiO2) synthesis...................................................................... 26

2.2.3. Characterization ........................................................................................................... 27

2.2.4. Reactivity studies ......................................................................................................... 30

2.3. Results and Discussion ....................................................................................................... 31

2.3.1. Synthesis and basic characterization ........................................................................... 31

2.3.2. Temperature-programmed desorption (TPD) of NH3, pyridine, and CO2 .................. 34

2.3.3. Infrared spectroscopy of pyridine ................................................................................ 37

2.3.4. Ethanol temperature-programmed desorption ............................................................. 40

2.3.5. Reactivity ..................................................................................................................... 44

2.4. Conclusions ........................................................................................................................ 57

Reference .................................................................................................................................... 59

ix

Synthesis of Transition Metal Phosphides via Phosphate and Phosphite Method and Chapter 3

their Activity in the Hydrodeoxygenation of 2-Methyltetrahydrofuran ....................................... 62

3.1. Introduction: ....................................................................................................................... 62

3.2. Materials and Experiments ................................................................................................. 64

3.2.1. Materials and catalysts................................................................................................. 64

3.2.2. Reactivity studies ......................................................................................................... 68

3.3. Results and Discussion ....................................................................................................... 70

3.3.1. Temperature-programmed reduction ........................................................................... 70

3.3.2. BET surface area ......................................................................................................... 75

3.3.3. CO chemisorption ........................................................................................................ 76

3.3.4. X-ray diffraction .......................................................................................................... 77

3.3.5. X-ray photoelectron spectroscopy ............................................................................... 80

3.3.6. Reactivity in hydrodeoxygenation (HDO) of 2-methyltetrahydrofuran (2-MTHF) .... 83

3.4. Conclusions ...................................................................................................................... 102

3.5. Supplementary Information.............................................................................................. 104

Reference ................................................................................................................................. 106

Kinetic and Spectroscopic Study of Deoxygenation of 2-Methyltetrahydrofuran over Chapter 4

Supported Nickel Phosphide Catalyst ......................................................................................... 108

4.1. Introduction ...................................................................................................................... 108

4.2. Materials and Experiments ............................................................................................... 108

4.2.1. Materials and synthesis .............................................................................................. 108

4.2.2. General characterization ............................................................................................ 109

4.2.3. Low temperature reactivity test ................................................................................. 109

4.2.4. Kinetic test ................................................................................................................. 111

4.2.5. Temperature-programmed desorption ....................................................................... 111

4.2.6. Fourier transform infrared (FTIR) spectroscopic experiments .................................. 112

4.3. Results and Discussion ..................................................................................................... 113

4.3.1. Characterization ......................................................................................................... 113

4.3.2. Low temperature reactivity test ................................................................................. 115

4.3.3. Kinetic study .............................................................................................................. 116

4.3.4. Temperature-programmed desorption ....................................................................... 121

x

4.3.5. Fourier transform spectroscopy ................................................................................. 122

4.4. Conclusions ...................................................................................................................... 133

4.5. Supplementary Information.............................................................................................. 133

4.5.1. Crystallite size calculations ....................................................................................... 133

4.5.2. Weisz – Prater criterion for internal diffusion ........................................................... 134

Reference ................................................................................................................................. 135

Conclusions ................................................................................................................ 136 Chapter 5

5.1. Conclusions ...................................................................................................................... 136

5.2. Recommendations for future work ................................................................................... 137

xi

Lists of Tables

Table 1.1. Summary of recent studies on catalytic HDO of dibenzofuran and benzofuran ........... 6

Table 1.2. Summary of recent studies on catalytic HDO of furan

O

...................................11

Table 1.3. Summary of studies on catalytic HDO of tetrahydrofuran

O

............................... 13

Table 2.1.The true width at half-maximum of the peak and crystallite size of Ni2P/SiO2 catalyst

....................................................................................................................................................... 33

Table 2.2. Characterization results for HZSM-5 and Ni2P/SiO2 catalysts ................................... 34

Table 2.3. Acid and base properties for HZSM-5 and Ni2P/SiO2 catalysts .................................. 39

Table 2.4. Distribution of ethylene and acetaldehyde in different temperature ranges on EtOH-

TPD ............................................................................................................................................... 41

Table 2.5. Parameters in the Gorte Criterion ................................................................................ 44

Table 2.6. Reactivity in ethanol deoxygenation (Contact time = 162 s) ...................................... 45

Table 2.7. Rate constants used in the simulation ......................................................................... 52

Table 2.8. Summary of FTIR assignments ................................................................................... 56

Table 3.1. Materials and amounts used in the preparation of phosphides supported on silica ..... 66

Table 3.2. Characterization results for the supported phosphide catalysts. (I) – catalysts made by

the phosphite method, (A) – catalysts made by the phosphate method ........................................ 76

Table 3.3. Spectral parameters for Ni2P/SiO2 and WP/SiO2 catalysts .......................................... 80

Table 3.4. Conversions and apparent activation energies of 2-MTHF HDO on transition metal

phosphides..................................................................................................................................... 84

Table 3.5. Product selectivity for HDO of 2-MTHF on transition metal phosphides at 275 oC

*Conversions of FeP/SiO2 were at 300oC ..................................................................................... 89

Table 3.6. Parameters in the Weisz-Prater Criterion ................................................................... 104

Table 3.7. The observed reaction rates and corresponding Weisz-Prater criterion for Ni2P/SiO2

..................................................................................................................................................... 104

Table 3.8. The observed reaction rates and corresponding Weisz-Prater criterion for WP/SiO2 (I)

..................................................................................................................................................... 105

Table 3.9. The observed reaction rates and corresponding Weisz-Prater criterion for WP/SiO2 (A)

..................................................................................................................................................... 105

Table 4.1: Masses monitored during temperature programmed desorption of 2-MTHF and their

relative intensity ...........................................................................................................................112

Table 4.2: Characterization results of Ni2P/SiO2; * BET surface area value of SiO2 ..................115

Table 4.3 Effects of H2 and 2-MTHF partial pressures on the conversion of 2-MTHF over

Ni2P/SiO2 .....................................................................................................................................117

Table 4.4: Assignment of FTIR peaks ......................................................................................... 132

Table 4.5. Parameters in the Weisz-Prater Criterion ................................................................... 134

Table 4.6: CWP values for reactivity test at low temperature ...................................................... 135

xii

List of Figures

Fig. 1.1. Triangular prism (mono-phosphide - MP) and tetrakaidecahedral structure (metal-rich

phosphide – M2P) [53] .................................................................................................................... 4

Fig. 2.1. Synthesis and XRD Analysis of Ni2P/SiO2. ................................................................... 32

Fig. 2.2. Mass 17 signal from NH3-TPD ...................................................................................... 35

Fig. 2.3. Mass 79 signal from pyridine-TPD ................................................................................ 36

Fig. 2.4.. Mass 44 signal from CO2-TPD ...................................................................................... 37

Fig. 2.5. Infrared spectra of adsorbed pyridine over samples at 423 K ........................................ 38

Fig. 2.6. Infrared spectra of adsorbed pyridine over samples at 423 K ........................................ 41

Fig. 2.7. Conversion and Selectivity from EtOH-TPD ................................................................. 42

Fig. 2.8. Variation of ethanol conversion and product yield as a function of reaction temperature

at a contact time of 162 s. a) HZSM-5 b) Ni2P/SiO2 ................................................................. 46

Fig. 2.9. Variation of ethanol conversion and product yield as a function of contact time on

HZSM-5. a) Reaction temperature of 473 K b) Reaction temperature of 498 K .......................... 47

Fig. 2.10. Variation of ethanol conversion and product yield as a function of contact time on

Ni2P/SiO2. The curves are fits from a simulation model discussed in the text. ............................ 49

Fig. 2.11. EtOH FTIR over Ni2P/SiO2 with 0.15% EtOH in He flow ......................................... 53

Fig. 2.12. EtOH FTIR over Ni2P/SiO2 in He at 497 K ................................................................. 55

Fig. 3.1. Temperature-programmed reduction of transition metal phosphide catalysts supported

on silica; heating rate of 2oC/min, H2 flow rate of 100ccm/g of catalyst. a) Mass 18 signal,

phosphite precursors, b) Mass 18 signal, phosphate precursors, c) Mass 34 (PH3) signal,

phosphite precursors, d) Mass 34 (PH3) signal, phosphate precursors. ........................................ 71

Fig. 3.2. X-ray diffraction patterns of supported phosphide catalysts .......................................... 78

Fig. 3.3. Ni 2p and P 2p core-level spectra for silica supported nickel phosphide catalysts ........ 81

Fig. 3.4. W 4f and P 2p core-level spectra for silica supported tungsten phosphide catalysts ...... 82

Fig. 3.5. Turnover frequency of 2-MTHF on transition metal phosphides ................................... 85

Fig. 3.6. Total conversion and selectivity towards HDO products of 2-MTHF reaction on

transition metal phosphides........................................................................................................... 86

Fig. 3.7. Product selectivity of 2-MTHF reaction on Ni2P/SiO2 and CoP/SiO2 at a total

conversion of around 5% .............................................................................................................. 89

Fig. 3.8. Product selectivity of 2-MTHF reaction on WP/SiO2 and MoP/SiO2 at total conversion

of 5% ............................................................................................................................................. 91

Fig. 3.9. Product selectivity of 2-MTHF reaction on FeP/SiO2 and Pd/Al2O3 at total conversion

of 5% ............................................................................................................................................. 93

Fig. 3.10. Contact time study on Ni2P/SiO2 .................................................................................. 95

Fig. 3.11. Contact time study on WP/SiO2 prepared by the phosphite and phosphate methods ... 98

xiii

Fig. 4.1 X-ray diffraction pattern of Ni2P/SiO2 ...........................................................................114

Fig. 4.2. Arrhenius plot of 2-MTHF conversion over Ni2P/SiO2 .................................................116

Fig. 4.3. Selectivity of products in partial pressure experiment ..................................................119

Fig. 4.4. Estimate of reaction order of H2 and 2-MTHF ............................................................. 120

Fig. 4.5: Temperature programmed desorption of 2-MTHF on Ni2P/SiO2 in H2 and He at

atmospheric pressure, heating rate of 3 oC min

-1, injection temperature was at 30

oC ............... 121

Fig. 4.6 Temperature programmed desorption of 2-MTHF on Ni2P/SiO2 in H2 at atmospheric

pressure, heating rate of 3 oC min

-1, injection temperature was at 150

oC ................................. 122

Fig. 4.7. Temperature programmed desorption of 2-MTHF from the support SiO2 in H2 ......... 124

Fig. 4.8 Proposed structural models of (a) P2O5/SiO2 and (b) Ni2P/SiO2 [8] ............................. 125

Fig. 4.9. Temperature programmed desorption of 2-MTHF from Ni2P/SiO2 in H2 .................... 127

Fig. 4.10 Temperature programmed desorption of 2-MTHF from Ni2P/SiO2 in He .................. 128

Fig. 4.11. FTIR spectra of 2-MTHF adsorbed on Ni2P/SiO2 in 0.32% 2-MTHF/H2 flow as a

function of temperature ............................................................................................................... 130

Fig. 4.12. FTIR spectra of 2-MTHF adsorbed on Ni2P/SiO2 in 2-MTHF/He flow as a function of

temperature ................................................................................................................................. 131

Fig. 4.13. Transient FTIR spectra following the introduction of 2-MTHF onto Ni2P/SiO2 in H2

..................................................................................................................................................... 132

List of Schemes

Scheme 1.1. Simplified reaction network for dibenzofuran ........................................................... 9

Scheme 1.2. Simplified reaction network for benzofuran .............................................................. 9

Scheme 1.3. Reaction network of furan on various catalysts ....................................................... 12

Scheme 1.4. Reaction network of tetrahydrofuran on CoMo/Al2O3 ............................................ 14

Scheme 1.5. Reaction network of tetrahydrofuran on supported Pt catalysts .............................. 15

Scheme 1.6. Reaction network of 2-methyltetrahydrofuran on Pt catalysts ................................. 16

Scheme 2.1. “Rake” mechanism for ethanol reaction ................................................................... 50

Scheme 3.1. HDO reaction network of 2-MTHF on Ni2P/SiO2 ................................................... 97

Scheme 3.2. HDO reaction network of 2-MTHF on WP/SiO2 (phosphate method) .................. 100

Scheme 3.3. HDO reaction network of 2-MTHF on WP/SiO2 (phosphite method) ................... 102

xiv

List of Abbreviation

BF Benzofuran

BET Brunauer, Emmett and Teller surface area

CUS Coordinatively Unsaturated Sites

DBF Dibenzofuran

FTIR Fourier Transform Infrared spectroscopy

HDN Hydrodenitrogenation

HDO Hydrodeoxygenation

HDS Hydrodesulfurization

HYD Hydrogenation

HMS Hybrid Mesoporous Silica

2-MTHF 2-Methyltetrahydrofuran

THF Tetrahydrofuran

TPD Temperature Controlled Desorption

TPR Temperature Controlled Reduction

XRD X-ray Diffraction

XPS X-ray Photoelectron Spectroscopy

1

Chapter 1

Introduction

1.1. Importance of Hydrodeoxygenation

The search for renewable energy sources has been receiving much attention for many

reasons such as depletion of fossil fuels, new stringent environmental regulations, and rapid

growth of global energy consumption [1]. Biomass has several advantages over conventional

fossil fuels as an energy source. For example, biomass has a net zero CO2 production, negligible

SOx and NOx emissions, and abundant availability [2-8]. Raw biomass materials come from

various sources, including wastes and by-products from agriculture and forestry [9-11].

Therefore, utilization of biomass as an energy source contributes greatly to the recycling industry

and the conservation of the environment. For these reasons, biomass presents one of the most

promising energy sources.

Biomass can be converted to liquid products directly through liquefaction processes at

high pressures [10, 12-14]. In fast and flash pyrolysis, biomass is heated rapidly (~ 300 oC min

-1

to up to 700 oC) at a relatively low pressure (up to 0.2 MPa) in the absence of oxygen and

produces the highest yield of liquid products (50-70 %) [10, 15, 16]. However, raw biomass

derived oil is not suitable for direct use as fuel. Typically, the raw oil contains 35-50 wt% of

oxygenated compounds in which water takes up to 30 wt% [10, 17-20]. The high oxygen content

results in many disadvantages such as immiscibility with hydrocarbon fuels, low heating value,

and low chemical and thermal stability [4, 21-23]. Therefore, removing oxygenated compounds

is essential in the treatment of bio-oil.

2

There are three typical ways to reduce oxygen content: hydrodeoxygenation (HDO),

decarbonylation, and catalytic cracking [24]. The HDO process entails cleavage of C-O bonds in

the presence of hydrogen to form hydrocarbons and environmentally benign water. This

resembles the hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) processes in

petroleum feedstock hydro-treatment. These processes have been studied in depth, creating an

excellent background for research in HDO [25]. Furthermore, the HDO process can be

incorporated to the existing petroleum refining infrastructure and technology, an important fact

that reinforces the potential of HDO as one of the most promising methods to upgrade bio-oil

[26-30].

1.2. Catalysts in Hydrodeoxygenation

Hundreds of oxygen-rich substances in bio-oil have been identified including guaiacol,

aldehydes, ketones, furans, organic acids, and phenolic compounds [3, 20, 31, 32]. Furimsky

provided a thorough review of catalytic hydrodeoxygenation [4]. Traditional hydrotreating

catalysts (such as CoMo/Al2O3 or NiMo/Al2O3) have been investigated for HDO reactivity on

bio-oil [31, 33-36] as well as bio-oil model compounds [37-42]. Supported precious metal

catalysts were also examined [43-47]. Presulfidation requirement of the traditional CoMo and

NiMo catalysts and high costs and scarcity of the noble metal catalysts are among the drawbacks

of these materials.

The choice of catalyst supports plays an important role in HDO activity. One of the most

popular catalyst supports is Al2O3; however, Al2O3 has been shown to be unsuitable for HDO. In

the presence of large amounts of water Al2O3 converts to boehmite [16, 48, 49]. The formation of

boehmite results in oxidation of nickel as well as blockage of Mo and Ni sites on the catalyst.

3

Moreover, Popov et al. suggested that the acidity of Al2O3 led to the high affinity for carbon

formation (coking) on this type of support [50]. Among alternative support materials, SiO2 has

been used due to its neutrality and relatively low affinity for coking. Other supports were also

investigated such as carbon, ZrO2, and CeO2 [43, 51, 52]. However, SiO2 still has a better

economic edge due to its lower cost and the ease of preparation and utilization.

1.2.1. Transition metal phosphides

Emerging alternatives to sulfur containing catalysts and noble metal catalysts are

transition metal carbides, nitrides and phosphides. These catalysts exhibit combined physical

properties of both metals and ceramic, and thus are good conductors of heat and electricity, are

hard and strong and have high thermal and chemical stability [53, 54]. Though having similar

physical properties, the phosphides differ from the carbides and nitrides substantially in their

crystal structure. The carbon in carbides and the nitrogen in nitrides have an atomic radius of

0.071 nm and 0.065 nm respectively, substantially smaller than that of the phosphorus (0.109

nm) in phosphides. Therefore, in the crystal structure of carbides and nitrides, the carbon and

nitrogen atoms reside in the interstitial spaces inside the octahedral spaces formed by the closed-

packed metal atoms, whereas the larger-sized phosphorus does not fit in these spaces and is

instead located in the center of triangular prisms of metal atoms (Fig. 1.1). The phosphides are

also different from the sulfides (present in the conventional molybdenum catalysts) as they do

not form flat layered structures; instead, they have more isotropic external morphologies that

give rise to globular shapes. As described later in the deoxygenation on traditional molybdenum

catalysts, many studies reported that the active sites are located at the edges of the layered MoS2

slabs (Section 1.4.1). Due to the globular shapes, the phosphides potentially permit greater

4

access to active corner and edge sites than the sulfides [55]. Transition metal carbides and

nitrides suffer deactivation in the presence of high sulfur level in petroleum feedstocks [56-58].

Meanwhile, transition metal phosphides possess enhanced stability, high sulfur resistance, and

high reactivity for hydrotreating reactions [53]. For these advantages, transition metal

phosphides are considered potential substitutes for conventional CoMo and NiMo catalysts [53,

55, 59]. Recent work has shown that MoP [60-62], WP [62-64], CoP [65-67], and Ni2P [53, 68-

72] are highly active for HDS and HDN. This knowledge serves as a background for further

investigation of their HDO activity.

Fig. 1.1. Triangular prism (mono-phosphide - MP) and tetrakaidecahedral structure

(metal-rich phosphide – M2P) [53]

1.2.2. Preparations of transition metal phosphides

Many methods were reported for the synthesis of transition metal phosphides such as

solid state reaction of metal and phosphorous, electrolysis of fused salts, and most popular

recently reduction of phosphate precursors [55, 73-76]. However, there is still much need for an

improved preparation method. Milder synthesis conditions as well as better carrier materials are

the main targets in order to reduce energy consumption during catalyst production and maintain a

high surface area for the catalysts. Metal phosphides are mostly supported on high surface area

materials to maximize their catalytic effects. Less acidic supports are preferred because they

favor both the formation of metal phosphides and the reduction of coke deposition [77, 78].

5

Precursors other than phosphates were also investigated [70, 79, 80]. Among the newer

precursors, metal phosphites stand out as an important development. Compared to phosphate,

phosphite contains phosphorus in a lower oxidation state and thus should be reducible at a lower

temperature [65, 70, 77, 81]. For instance, active Ni2P phase from phosphite precursor was

formed at a reduction temperature of 300 oC, whereas when using phosphate as precursor, the

active phase was formed at the reduction temperature of 500 oC [59, 82]. Conditions to prepare

phosphite precursors are also much less intense than those of phosphates. Specifically, phosphite

precursor preparation only requires drying at low temperatures (< 120 oC) and bypasses the

calcination at high temperatures (> 450 oC) that is mandated in phosphate preparation.

1.3. Hydrodeoxygenation of Cyclic Ether Compounds

A plethora of work has been conducted on the HDO of a wide range of oxygenated

compounds in which lignin derivatives (phenols and substituted phenols) received the most

attention [83-88]. The apparent reactivity of different compounds followed the order of: alcohol

> ketone > alkylether > carboxylic acid ~ m-/p- phenol ~ naphthol > phenol > diarylether ~ o-

phenol ~ alkylfuran > benzofuran > dibenzofuran [4]. Aromatic model compounds (phenols and

benzofurans) constitute the hardest classes of compounds to deoxygenate and therefore are often

chosen in HDO studies of model compounds. To overcome the high energy barrier of the

aromatic group, the hydrodeoxygenation process typically involves with hydrogenation (HYD)

of the unsaturated ring, followed by deoxygenation. Transition metal phosphides with good

hydrogen-transfer properties are well-capable to catalyze the initial hydrogenation and therefore

are excellent potential catalysts for hydrodeoxygenation.

6

1.3.1. Aromatic cyclic ethers

Cyclic ethers pose interesting topics for HDO investigations. Aromatic cyclic ethers, such

as dibenzofuran (DBF)

O

, benzofuran (BF)

O

, and furan

O

,

present in conventional petroleum feedstock (~0.1 wt%), heavy oil distillates (~2 wt%), oil shale

(~1 wt%), and coal derived liquids (~7 wt%) [4, 89, 90] and therefore receive considerable

coverage [4]. Though constituting a small amount, these compounds are important for their

resistance to HDO and their effects on the hydrodesulfurization (HDS) and the

hydrodenitrogenation (HDN) during hydrotreating processes [91]. A review by Furimsky

comprehensively covers various aspects of catalytic HDO of these compounds including

mechanism, kinetics, and interaction modes of the reactants with the catalyst surfaces [4]. More

recent studies are summarized in Table 1.1 and Table 1.2.

1.3.1.a) Hydrodeoxygenation of dibenzofuran (DBF) and benzofuran (BF)

Table 1.1. Summary of recent studies on catalytic HDO of dibenzofuran and benzofuran

Catalysts Remarks Reference

Various A review of studies prior to 2000 covered catalytic

hydrodeoxygenation of mostly conventional catalysts –

NiMo or CoMo on alumina.

Carbides and nitrides showed to be less active than

conventional catalysts.

Furimsky

[4],

Ramanathan

[92]

Studies on Dibenzofuran (2000-2012)

O

Pt/ mesoporous ZSM5

Pt/ ZSM5

Reaction network: hydrogenation of one aromatic ring

occurred first.

Wang [93]

7

Pt/Al2O3 Mesoporous ZSM-5 support enhanced the catalyst

activity compared to the alumina support.

Studies on Benzofuran (2000-2012)

O

Sulfided NiMoP/Al2O3 Reaction network: partial hydrogenation of the furan

ring occurred first.

Self-inhibition of BF and intermediate oxygenated

species was observed.

Presence of H2S promoted HDO and hydrogenation

activity.

Romero

[83]

Mo/Al2O3

NiMo/Al2O3 (sulfided

and reduced)

Reaction network: partial hydrogenation of the furan

ring occurred first.

The Ni promoter promoted hydrogenation and HDO.

The reduced catalyst was superior to the sulfided

catalyst.

The reduced catalyst contained coordinatively

unsaturated sites and favored full hydrogenation of 2,3-

dihydrobenzofuran prior to oxygen removal.

The sulfided catalyst contained Brönsted acid sites and

favored hydrogenolysis of 2,3-dihydrobenzofuran to

ethylphenol prior to oxygen removal.

The presence of H2S decreased HDO activity.

Bunch [38,

39, 94]

Pt/activated carbon in

super critical water

Reaction network: partial hydrogenation of the furan

ring occurred first.

Self-inhibition of BF and intermediate oxygenated

species was observed.

Dickinson

[95]

Pt/ SiO2-Al2O3

Pd/ SiO2-Al2O3

PtxPdy/ SiO2-Al2O3

HDO reactivity: PtPd (1:4) alloy > Pd > Pt.

Reaction network: full hydrogenation of the aromatic

ring occurred prior to oxygen removal.

Self-inhibition of BF and intermediate oxygenated

species was observed.

Liu [96]

Ni2P/SiO2 Catalyst loading amount was optimum at 18.1wt% with

HDO conversion of 80%.

Low loading amount led to incomplete formation of

Ni2P phase while high loading amount led to formation

of Ni-Ni bond which is susceptible to coking.

Oyama [68]

On conventional NiMo or CoMo catalysts, HDO of DBF and BF exhibits hydrogenation,

hydrogenolysis and direct deoxygenation. The main products were substituted phenols, aromatic

and cyclic hydrocarbons (Scheme 1.1,Scheme 1.2). Self-inhibition by intermediate products and

sulfur containing compounds as well as poisoning by nitrogen containing compounds was

8

observed [97-99]. The mutual inhibition of HDO and HDS indicates competitive adsorptions at

the same active sites. This was supported by the oxygen-chemisorption results that showed linear

correlation with HDS activity [100] and the same adverse effect of coking on HDS and HDO

[101]. The HDO of DBF and BF showed products from both hydrogenation and hydrogenolysis

routes (Scheme 1.1, Scheme 1.2). The existence of different active sites for hydrogenation and

hydrogenolysis has been frequently proposed in the literature [102] though both types of active

sites involve coordinatively unsaturated Mo surface atoms (CUS) formed by sulfur vacancies

associated with Mo atoms at the edges of MoS2 slabs [83, 94]. The distinction between

hydrogenation and hydrogenolysis sites was found related to the degree of uncoordination [103,

104], the Brønsted-Lewis acidic character [102, 105], or the environment of the vacancy [106-

108]. A small amount of H2S in the feed was reported to be beneficial to the overall HDO as it

maintained the MoS2 slabs while some portion of surface sulfur was replaced with oxygen atoms

from the reactant [109, 110].

Laurent and Delmon proposed that C-O hydrogenolysis of DBF could be based on an

ensemble of coordinatively unsaturated Mo atoms and the hydrogenation (HYD) sites located on

one triply unsaturated Mo atom [40]. The HYD sites may have been responsible for forming a π-

interaction with one of the aromatic rings and weaken the C-O bond, leading to hydrogenolysis.

The HYD sites also gave rise to partially or fully hydrogenated products. Direct deoxygenation

from DBF may originate from σ-bonded adsorption of the oxygen to the hydrogenolysis site.

Direct deoxygenation of DBF is not as straightforward as direct desulfurization of DBT because

oxygen is less polarizable than sulfur [111].

Similar deductions could be drawn for BF in which the π-bond with the HYD sites results

in a partial hydrogenation of BF to 2,3-dihydroBF, then interactions with hydrogenolysis sites

9

cleave C-O bond to produce ethylphenol. Unlike DBF, the direct deoxygenation of BF was not

observed experimentally on the conventional NiMo or CoMo catalysts.

O

O

OH

O OH

Single-ring hydrocarbons

Double-ring hydrocarbonsDibenzofuran (DBF)

Deoxygenation

Deoxygenation

1

2

1 2

1. Hydrogenation (HYD)

2. Hydrogenolysis

Scheme 1.1. Simplified reaction network for dibenzofuran

O O OH

CH3 CH3

CH3

OH

CH3

Various pathways

phenols, benzene, cyclic hydrocarbons, etc.

Benzofuran (BF)

1. Hydrogenation (HYD)

2. Hydrogenolysis

1 2

1 & 2

1 & 2

Scheme 1.2. Simplified reaction network for benzofuran

There are few studies on HDO of DBF for catalysts other than conventional nickel or

cobalt molybdenum catalysts. Most recently, noble metal (Pt) on various supports were tested

[93]. Mesoporous hierarchical ZSM5 zeolites have combined advantages of uniform meso-pore

size and strong acidity and were expected to improve catalytic activities. The catalyst indeed was

better than the supported alumina and microporous ZSM5. Considering that the supports

(mesoporous ZSM5, ZSM5 and alumina) have plenty of acidic sites and that acidic sites are

10

generally considered hydrogenolysis or dehydration sites, it is interesting that direct

deoxygenation did not occur on any of these catalysts. Instead, the catalysts showed strong

hydrogenation capacity and some dehydration activity. This indicated hydrogen spill-over effect

from the metal sites to the acidic sites, a well-known phenomenon for noble metals [102, 112].

A similar study for HDO of BF was conducted on Pt, Pd and alloyed PtPd supported on

silica-alumina [96]. Complete ring hydrogenation prior to deoxygenation was observed,

consistent with the hydrogenation capacity of metallic catalysts. The HDO activity followed the

order of PtPd (1:4) > Pd > Pt. On Pt/activated carbon, the major pathway involved initial partial

hydrogenation of the furan ring prior to hydrogenolysis [95]. The experiment was conducted in

supercritical water and increasing water loading led to a decrease in fully hydrogenated

compounds as well as total deoxygenation. This suggests that water may have inhibited

hydrogenation by strong bonding to HYD sites. Self-inhibition from the reactant and the

oxygenated intermediates was also observed, consistent with previous studies on other catalysts

[94, 99].

1.3.1.b) Hydrodeoxygenation of furan

The simplest aromatic oxygenated compound – furan – was studied mostly on traditional

catalysts consisting of promoted and unpromoted molybdenum on various supports and a few

other catalysts (Table 1.2). All experiments on furan were conducted at atmospheric pressure to

prevent complete hydrogenation to tetrahydrofuran (THF). The goal was to observe the

interactions between the furan ring and the catalyst.

11

Table 1.2. Summary of recent studies on catalytic HDO of furan

O

Catalysts Remarks Reference

Various Review (studies prior to 2000): Catalytic

hydrodeoxygenation of mostly conventional NiMo or

CoMo on alumina catalysts

Furimsky [4]

Studies on Furan (2000-2012)

Mo, NiMo, CoMo on

various supports (Al-

HMS, MCM41)

HMS = hybrid

mesoporous silica

HDO reactivity:

CoMo/Al-HMS > NiMo/Al-HMS > Mo/Al-HMS

NiMo/MCM41 > CoMo/MCM41 > Mo/MCM41

Chiranjeevi

[113-115]

MoS2

Density functional

theory (DFT) study

The stability of MoS2 catalyst depends on H2S/H2O

ratio during the reaction. The metallic edge is stable

regardless of the ratio; the sulfur edge needs a presence

of H2S to prevent partial oxygenation.

Furan adsorbs on coordinatively unsaturated sites

(vacancy of an S atom from the metallic edge).

Reaction energies are 0.5 eV for the η1 adsorption

mode of furan and 1.39 eV for the η5 adsorption mode.

Badawi [109]

Ru2P, RuP, Ru, and Ni2P

supported on SiO2,

CoMo/Al2O3

HDO activity:

Ru2P/SiO2 >> RuP/SiO2 > Ru/SiO2 > Ni2P/SiO2 >>

CoMo/Al2O3

Ru/SiO2 strongly favored C3 hydrocarbons while others

favored C4 hydrocarbons. Phosphorous may have increased

Lewis acid acidity, resulting in stabilization of a ring-

opened species having an η1(O) geometry. Meanwhile,

Ru/SiO2 contained mostly η2(C,O) species that favored

decarbonylation.

Bowker

[116]

Furan hydrodeoxygenation occurs via direct deoxygenation, hydrogenation,

hydrogenolysis, and cracking with C3 and C4 hydrocarbons as major products (Scheme 1.3). In

all of the studies (Table 1.2), no dihydrofuran or tetrahydrofuran in the product stream was

observed indicating that initial partial hydrogenation of the ring was fast and was followed

immediately by a ring opening (Scheme 1.3). Promoted catalysts performed better than

unpromoted catalysts and this was shown on both molybdenum and tungsten catalysts on various

12

supports [113-115, 117, 118]. The promoter did not increase the number of active sites, rather it

promoted the intrinsic activity of the sites.

Scheme 1.3. Reaction network of furan on various catalysts [116]

The simplicity of furan allowed detailed density functional theory calculations [116]. On

MoS2, the calculations showed furan only adsorbs onto defective metal sites that have at least

one anion sulfur vacancy. The calculations also showed that a small presence of H2S is needed in

order to keep the MoS2 from being oxygenated, this is consistent with a previous study [110].

1.3.2. Saturated cyclic ethers

Under a high H2 pressure of typical hydrotreating conditions, unsaturated cyclic ethers

are partially or fully hydrogenated. Thus, saturated cyclic ethers are an important intermediate

before the final oxygen removal. Hydrodeoxygenation of tetrahydrofuran has been investigated

on CoMo/Al2O3, supported Pt, and supported Mo catalysts (Table 1.3)

13

1.3.2.a) Hydrodeoxygenation of tetrahydrofuran (THF)

Table 1.3. Summary of studies on catalytic HDO of tetrahydrofuran

O

Catalysts Remarks Reference

CoMo/Al2O3 HDO products are similar to furan HDO (butenes,

propene, etc.)

Sulfided catalyst provided more efficient transfer of

surface hydrogen to the adsorbed THF thus increased

HDO activity and suppressed decarbonylation

compared to non-sulfided catalyst.

Little HDO reactivity on the Al2O3 support produced

same products with the reduced catalyst.

Furimsky

[119]

Pt on TiO2, SiO2, and

Al2O3

HDO activity:

Pt/TiO2 > Pt/SiO2 and Pt/Al2O3 >> Pt

Cracking formed CO which poisoned the catalyst. At

high temperature (over 200 oC) the adsorbed CO was

reformed into methane.

C-O scission occurred on the Pt/support phase

boundary.

Kreuzer

[120],

Bartok [121]

Mo on TiO2, ZrO2, and

TiO2-ZrO2

HDO activity:

Mo/TiO2-ZrO2 > Mo/TiO2 > Mo/ZrO2

Mixed oxide supports may have facilitated formation

of coordinatively unsaturated sites, increasing HDO

activity.

Maity [122]

Over CoMo/Al2O3 catalyst, HDO of THF produced dehydrogenated products including

butenes (sulfided catalysts) and propylene and butadiene (reduced catalyst) (Scheme 1.4). More

hydrogenated products on the sulfided catalyst suggested a better surface hydrogen transfer than

on the reduced catalyst. About 39% of THF converted to tar and coke on the reduced catalyst

(compared to 8% for the sulfided catalyst). The lack of effective hydrogen surface species may

have caused polymerization of butadienes. No alcohol or aldehyde products were observed,

indicating the ring opened intermediate stayed attached to the catalyst surface via the O atom

throughout the deoxygenation. In the absence of hydrogen, there was a small extent of HDO

14

activity over the reduced catalysts as well as over the Al2O3 support suggesting intramolecular

migration of THF’s hydrogen atoms.

O+

*Tetrahydrofuran

O

* CO+

1-Butene

Propene

Butadiene

Scheme 1.4. Reaction network of tetrahydrofuran on CoMo/Al2O3

Over noble metal catalysts, HDO of THF produced fully hydrogenated products such as

butane, propane, and 1-butanol (Scheme 1.5). The hydrogenation preference on metal catalysts is

consistent with the studies on BF and DBF (Table 1.1). Both Kreuzer et al. [120] and Bartok et

al. [121] investigated the HDO of THF on the Pt catalyst and obtained similar products, yet they

proposed different reaction pathways. Kreuzer et al. proposed that 1-butanol was the first ring

opened product which could go through a hydrogenolysis to give butane or transform into

butanal – a secondary product that reacted to form propane and CO. Bartok et al. proposed a

sequence of surface reactions in which the formation of both 1-butanol and butane was from an

η1(O) surface butoxy species and the formation of butanal and propane was from an η2

(CO)

surface species. The supports showed a strong effect on the selectivity. The phase boundary of

the metal and the support provided active sites for C-O bond scission. The TiO2 supported Pt

yielded mostly butane while SiO2 and Al2O3 supported ones produced more decarbonylation

products and CO. In all cases, CO caused self-poisoning of the Pt surface.

15

O+

*

O

*

OH

O

**

O

Tetrahydrofuran

1-Butanol

+ CO

ButaneButanal

Propane

Scheme 1.5. Reaction network of tetrahydrofuran on supported Pt catalysts

1.3.2.b) Hydrodeoxygenation of 2-methyltetrahydrofuran (2-MTHF)

Research on HDO of 2-MTHF is limited. Platinum, palladium and nickel catalysts were

tested [123-126] but the most detailed results were for Pt catalysts. A proposed reaction network

for 2-MTHF involved the reactant adsorbed onto two active sites and proceeded in two pathways

(Scheme 1.6). The main pathway (A) occurred as the bond between oxygen and the primary

carbon atom was cleaved, producing 2-pentanone and 2-pentanol as primary product, then

pentene and pentane as secondary and final products. The pathway (B) involved scission

between the oxygen and the secondary carbon which led to 1-pentanol and pentanal as primary

products (observed only at 150 oC) and to pentane or decarbonylated to butane as final products.

The selectivity towards pathway (B) was 30% on the supported Pt and 7% on the unsupported Pt.

The silica support had no reactivity therefore the pathway (B) activity was due to the phase

boundary.

16

O+

**

O+

*2-Methyltetrahydrofuran

O+

* *

OH

CO

O

OH

O +

A

B

Scheme 1.6. Reaction network of 2-methyltetrahydrofuran on Pt catalysts

1.4. Conclusions

Commercialization of second generation biomass derived oil depends on an efficient and

economical oxygen removal system. The review on HDO of cyclic ethers emphasized the

importance of surface hydrogen species in the HDO reactivity. The traditional hydrotreating

catalysts (supported CoMo, NiMo, or Mo) require the presence of H2S – a well-known pollutant

– to maintain the stability of the active phase; yet, overdosing H2S causes adverse effect on the

HDO reactivity. Metal catalysts tend to be poisoned by CO produced from cracking reactions.

Transition metal phosphides were shown to have good hydrogen transfer capacity and good

resistance to sulfur and nitrogen contained compounds; therefore these catalysts have great

potential in hydrodeoxygenation applications.

1.5. Goals

17

The research presented in this dissertation aims to characterize supported transition metal

phosphides with a special focus on Ni2P/SiO2, evaluate their activity in HDO reactions, and

investigate the mechanisms of HDO. In order to reach the goal, the following tasks have been

carried out:

• Synthesis of transition metal phosphides on high surface area support via two methods –

phosphate method and phosphite method.

• Characterization of the catalysts with CO chemisorption, BET surface area

measurements, temperature-programmed desorption, and X-ray diffraction

measurements.

• NH3 temperature-programmed desorption on Ni2P/SiO2 to quantify the catalyst acidity.

• Fourier transformed infrared measurements to investigate the acidic properties of Ni2P

catalysts with pyridine as a probe molecule.

• Steady state reactivity and transient study of ethanol reactions on Ni2P/SiO2.

• Fourier transform infrared measurements of ethanol reactions on the Ni2P/SiO2 surface.

• Steady state reactivity and transient study of 2-methyltetrahydrofuran reactions on

Ni2P/SiO2.

• Fourier transform infrared measurements of 2-MTHF reactions on the Ni2P/SiO2 surface.

1.6. Dissertation Overview

Chapter 1 describes the motivation of the project and provides background information

including reviews on catalysts in hydrodeoxygenation processes and studies on catalytic

reactions of cyclic ethers.

18

Chapter 2 presents a study of ethanol decomposition on Ni2P/SiO2 and HZSM5. A

reaction network was proposed on Ni2P/SiO2.

Chapter 3 presents a study of 2-methyltetrahydrofuran reactions on a series of transition

metal phosphides. The phosphides were synthesized via two methods: the phosphate precursor

method and the phosphite precursor method. Reaction networks were proposed for the iron group

phosphides (Ni2P and CoP) and group 8 metal phosphides (WP and MoP).

Chapter 4 presents a Fourier transform infrared spectroscopic study of 2-

methyltetrahydrofuran on Ni2P/SiO2.

Chapter 5 presents the conclusions of this research and suggestions for future work.

19

References

1. Demirbas, A., Appl. Energ., 86 (2009) S108.

2. A. A. Lappas,S. Bezergianni,I. A. Vasalos, Catal. Today, 145 (2009) 55.

3. D. Mohan,C. U. Pittman Jr.,P. H. Steele, Energ. Fuel, 20 (2006) 848.

4. Furimsky, E., Appl. Catal. A: Gen., 199 (2000) 147.

5. J. I. C. Shumaker,S. A. Crofcheck,E. S. Tackett,M. Jiménez, Catal. Lett., 115 (2007) 56.

6. Klass, D. L., Biomass for renewable energy. Fuels and chemicals. 1998, San Diego:

Academic Press.

7. L. R. Lynd,J. H. Cushman,R. J. Nichols,C. E. Wyman, Science, 251 (1991) 1318.

8. S. Czernik, A.V. Bridgwater, Energ. Fuel, 18 (2004) 590.

9. M. F. Demirbas ,M. Balat, J. Sci. Ind. Res., 66 (2007) 797.

10. H. B. Goyal,D. Seal,R. C. Saxena, Renewable and Sustainable Energy Reviews, 12

(2008) 504.

11. McKendry, P., Bioresour. Technol., 83 (2002) 37.

12. Demirbas, A., Energ. Convers. Manage, 42 (2001) 1357.

13. Demirbas, A., Energ. Convers. Manage, 41 (2000) 633.

14. L. Zhang,C. Xu,P. Champagne, Energ. Convers. Manage., 51 (2010) 969.

15. A. V. Bridgwater ,G. V. C. Peacocke, Renewable Sustainable Energy Rev., 4 (2000) 1.

16. Elliott, D.C., Energ. Fuel, 21 (2007) 1792.

17. C. A. Mullen ,A. A. Boateng, Energ. Fuel, 22 (2008) 2104.

18. M. Bertero,G. de la Puente,U. Sedran, Fuel, 95 (2012) 263.

19. Bridgwater, T., Biomass Bioenergy, 31 (2007) 7.

20. J. H. Marsman,J. Wildschut,P. Evers,S. de Koning,H. J. Heeres, J. Chromatogr. A., 1188

(2008) 17.

21. A. Oasmaa ,S. Czernik, Appl. Catal. A: Gen., 13 (1999) 914.

22. Yaman, S., Energ. Convers. Manage., 45 (2004) 651.

23. Demirbas, M. F., Appl. Energ., 86 (2009) S151.

24. S. R. A. Kersten, W. P. M. van Swaaij, L. Lefferts, K. Seshan: G. Centi, R. A. van Saten

(Eds.), Catalysis for renewable - from feedstocks to energy production. 2007, Wienheim:

Wiley- VCH. Chapter 6.

25. R. Prins, M. Egorova, A. Rothlisberger, Y. Zhao, N. Sivasankar, P. Kukul, Catal. Today,

111 (2006) 84.

26. J. Sadhukhan ,K. S. Ng, Ind. Eng. Chem. Res., 50 (2011) 6794.

27. F. M. Mercader,M. J. Groeneveld,S. R. A. Kersten,N. W. J. Way, Appl. Catal. B:

Environ. , 96 (2010) 57.

28. J. A. Melero,A. Garcia,J. Iglesias, Advances in clean hydrocarbon fuel processing:

Science and technology. 2011, Cambridge: Woodhead Publishing. 199.

29. S. Fernando, S. Adhikari, C. Chandrapal, N. Murali Energ. Fuel, 20 (2006) 1727.

30. F. de M. Mercader,M. J. Groeneveld,S. R. A. Kersten,C. Geantet,G. Toussaint,N. W. J.

Way,C. J. Schaverien,K. J. A. Hogendoorn, Energy Environ. Sci., 4 (2011 ) 985.

31. P. M. Mortensen,J. D. Grunwaldt,P. A. Jensen,K. G. Knudsen,A. D. Jensen, Appl. Catal.

A: Gen., 407 (2011) 1.

32. G. W. Huber,S. Iborra,A. Corma, Chem. Rev., 106 (2006) 4044.

33. S. D. S. Murti, K. H. Choi, K. Sakanishi, O. Okuma, Y. Korai, I. Mochida, Fuel, 84

(2005) 135.

20

34. J. Wildschut,F. H. Mahfud,R. H. Venderbosch,H. J. Heeres, Ind. Eng. Chem. Res. , 48

(2009) 10324.

35. J. Zakzeski,P.C.A. Bruijnincx,A.L. Jongerius,B.M. Weckhuysen, Chem. Rev., 110 (2010)

3552.

36. V. N. Bui, G. Toussaint, D. Laurenti, C. Mirodatos, C. Geantet, Catal. Today, 143 (2009)

172.

37. A. Popov,E. Kondratieva,J. Gilson,L. Mariey,A. Travert,F. Mauge, Catal. Today, 172

(2011) 132.

38. A. Y. Bunch ,U. S. Ozkan, J. Catal. , 206 (2002) 177.

39. A. Y. Bunch,X. Wang,U. S. Ozkan, J. Mol. Catal. A, 270 (2007) 264.

40. E. Laurent ,B. Delmon, Appl. Catal. A: Gen., 109 (1994) 77.

41. E. M. Ryymin,M. L. Honkela,T. R. Viljava,A. O. Krause, Appl. Catal. A: Gen., 389

(2010) 114.

42. V. N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B: Environ., 101 (2011)

239.

43. A. Gutierrez,R. K. Kaila,M. L. Honkela,R. Slioor,A. O. I. Krause, Catal. Today, 147

(2009) 239.

44. D. Prochazkova,P. Zamostny,M. Bejblova,L. Cerveny,J. Cejka, Appl. Catal. A: Gen., 332

(2007) 56.

45. R. C. Runnebaum, R. J. Lobo-Lapidus, T. Nimmanwudipong, D. E. Block, B. C. Gates,

Energ. Fuel, 25 (2011) 4776.

46. T. Nimmanwudipong, R. C. Runnebaum, D. E. Block, B. C. Gates, Energ. Fuel, 25

(2011) 3417.

47. T. T. Pham, L. L. Lobban, D. E. Resasco, R. G. Mallinson, J. Catal., 206 (2009) 9.

48. E. Laurent,A. Centeno,B. Delmon. in 6th Inter. Symp. Catalyst Deactivation. 1994.

49. R. H. Venderbosch, A. R. Ardiyanti, J. Wildschut, A. Oasmaa, H. J. Heeres, J. Chem.

Technol. Biotechnol. , 85 (2010) 674.

50. A. Popov,E. Kondratieva,J. M. Goupil,L. Mariey,P. Bazin,J. -P. Gilson,A. Travert,F.

Mauge, J. Phys. Chem. C, 114 (2010) 15661.

51. V. A. Yakovlev, S. A. Khromova, O. V. Sherstyuk, V. O. Dundich, D. Y. Ermakov, V.

M. Novopashina, M. Y. Lebedev, O. Bulavchenko, V. N. Parmon, Catal. Today, 144

(2009) 362.

52. V. N. Bui, D. Laurenti, P. Delichère, C. Geantet, Appl. Catal. B: Environ., 111 (2011)

246.

53. S. T. Oyama, T. Gott, H. Zhao, Y. -K.Lee, Catal. Today, 143 (2009) 94.

54. Oyama, S. T., Transition metal carbides, nitrides, and phosphides. Handbook of catalysis,

ed. H.K. G. Ertl, J. Weitkamp. 2008, Weinheim: Springer-Verlag.

55. Oyama, S. T., J. Catal., 216 (2003) 343.

56. B. Dhandapani,T. St. Clair,S. T. Oyama, Appl. Catal. A: Gen, 168 (1998) 219.

57. B. Diaz,S. J. Sawhill,D. H. Bale,R. Main,D. C. Phillips,S. Korlann,R. Self,M. E. Bussell,

Catal. Today, 86 (2003) 191.

58. Furimsky, E., Appl. Catal. A: Gen, 240 (2003) 1.

59. S. T. Oyama, X. Wang, Y. -K. Lee, K. Bando, F. G. Requejo, J. Catal., 210 (2002) 207.

60. P. Clark,X. Wang,S. T. Oyama, J. Catal., 207 (2002) 256.

61. W. Li,B. Dhandapani,S. T. Oyama, Chem. Lett. , (1998) 207.

62. P. Clark,X. Wang,P. Deck,S. T. Oyama, J. Catal., 210 (2002) 116.

21

63. P. Clark,W. Li,S. T. Oyama, J. Catal., 200 (2001) 140.

64. S. T. Oyama, P. Clark, X. Wang, T. Shido, Y. Iwasawa, S. Hayashi, J. M. Ramallo-Lopez,

F. G. Requejo, J. Phys. Chem. B, 106 (2002) 1913.

65. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, Appl. Catal.

B: Environ., 92 (2009) 100.

66. X. Wang,P. Clark,S. T. Oyama, J. Catal., 208 (2002) 321.

67. V. Zuzaniuk ,R. Prins, J. Catal., 219 (2003) 85.

68. S. T. Oyama, X. Wang, Y. -K. Lee, W.- J. Chun, J. Catal., 221 (2004) 263.

69. S. T. Oyama, Y. –K. Lee, J. Catal., 258 (2008) 393.

70. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, J. Phys.

Chem. C, 113 (2009) 17032.

71. C. Stinner,R. Prins,Th. Weber, J. Catal., 202 (2001) 187.

72. Y. Shu ,S. T. Oyama, Carbon, 43 (2005) 1517.

73. Chene, M., Anal. Chim. , 15 (1941) 187.

74. West, A. R., Solid state chemistry and its applications. 1992, Chichester: Wiley.

75. S. Burns J. S. J. Hargreaves, S. M. Hunter, Catal. Comm. , 8 (2007) 931.

76. A. Montesinos-Castellanos,T. A. Zepeda,B. Pawelec,J. L. G. Fierro,J. A. de los Reyes,

Chem. Mater. , 19 (2007) 5627.

77. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, J. Catal., 263

(2009) 4.

78. T. I. Korányi, Z. Vít, D. G. Poduval, R. Ryoo, H. -S. Kim, E. J. M. Hensen, J. Catal., 253

(2008) 119.

79. H. Loboue,C. Guillot-Deudon,A. Florin Popa,A. Lafond,B. Rebours,C. Pichon,T.

Cseri,G. Berhault,C. Geantet, Catal. Today, 130 (2008) 63.

80. Z. W. Yao,L. Wang,H. Dong, J. Alloy Compd., 473 (2009) L10.

81. Q. Guan, W. Li, M. Zhang, K. Tao, J. Catal., 263 (2009) 1.

82. G. Shi ,J. Shen, J. Mater. Chem. , 19 (2009) 2295.

83. Y. Romero,F. Richard,Y. Reneme,S. Brunet, Appl. Catal. A: Gen., 353 (2009) 45.

84. M. Badawi,J. F. Paul,S. Cristol,E. Payen,Y. Romero,F. Richard,S. Brunet,D. Lambert,X.

Portier,A. Popov,E. Kondratieva,J. M. Goupil,J. El Fallah,J. P. Gilson,L. Mariey,A.

Travert,F. Mauge, J. Catal., 282 (2011) 155.

85. H. Y. Zhao,D. Li,P. Bui,S. T. Oyama, Appl. Catal. A: Gen., 391 (2011) 305.

86. C. Zhao,Y. Kou,A. A. Lemonidou,X. Li,J. A. Lercher, Angew. Chem. Int. Ed. , 48 (2009)

3987.

87. F. E. Massoth,P. Politzer,M. C. Concha,J. S. Murray,J. Jakowski,J. Simons, J. Phys.

Chem. B, 110 (2006) 14283.

88. N. Yang,Y. Yuan,R. Dykeman,Y. Kou,P. Dyson, Angew. Chem. Int. Ed., 49 (2010) 5549.

89. Furimsky, E., Catal. Rev. Sci. Eng. , 25 (1983) 421.

90. Leckel, D., Energ. Fuel, 22 (2008) 231.

91. C. Lee ,D. F. Ollis, J. Catal., 87 (1984) 332.

92. S. Ramanathan, S. T. Oyama, J. Phys. Chem., 99 (1995) 16365.

93. Y. Wang,Y. Fang,T. He,H. Hu,J. Wu, Catal. Comm., 12 (2011) 1201.

94. A. Y. Bunch,X. Wang,U. S. Ozkan, Appl. Catal. A: Gen., 346 (2008) 96.

95. J. G. Dickinson,J. T. Poberezny,P. E. Savage, Appl. Catal. B: Environ., 123-124 (2012)

357.

96. C. Liu,Z. Shao,Z. Xiao,C. T. Williams,C. Liang, Energ. Fuel, 26 (2012) 4205.

22

97. C. Kordulis,A. Gouromihou,A. Lycourghiotis,C. Papadopoulo,H. K. Mataralis, Appl.

Catal., 67 (1990) 39.

98. V. LaVopa, C. N. Satterfield, J. Catal., 110 (1998) 375.

99. E. O. Odebunmi ,D. F. Ollis, J. Catal., 80 (1983) 65.

100. S. J. Tauster, T. A. Pecorara, R. R. Chianelli, J. Catal., 63 (1980) 515.

101. Y. Yamamoto,F. Kumata,F. E. Massoth, Fuel Proc. Technol., 19 (1998) 253.

102. H. Du,C. Fairbridge,H. Yang,Z. Ring, Appl. Catal. A: Gen., 294 (2005) 1.

103. Delmon, B., Bull. Soc. Chim. Belg., 88 (1979) 979.

104. Y. Okamoto,A. Maezawa,T. Imanaka, J. Catal., 120 (1989) 29.

105. C. Moreau,C. Aubert,R. Durand,N. Zmimita,P. Geneste, Catal. Today, 4 (1988) 117.

106. B. Delmon ,J. L. Dallons, Bull. Soc. Chim. Belg., 97 (1988) 473.

107. J. Miciukiewicz,W. Zmierczak,F. E. Massoth, Bull. Soc. Chim. Belg., 96 (1987) 915.

108. V. Stuchly, L. Beranek, Appl. Catal., 35 (1987) 35.

109. M. Badawi,S. Cristol,J. Paul,E. Payen, C. R. Chimie, 12 (2009) 754.

110. T. Kabe, W. Quian, A. Ishihara, J. Phys. Chem., 98 (1994) 912.

111. M. J. Girgis ,B. C. Gates, Ind. Eng. Chem. Res., 33 (1994) 1098.

112. U. Roland, T. Braunschweig, F. Roessner, J. Mol. Catal. A, 127 (1997) 61.

113. T. Chiranjeevi, G. M. Kumaran, J. K. Gupta, G. M. Dhar, Catal. Comm., 6 (2005) 101.

114. T. Chiranjeevi, G. M. Kumaran, J. K. Gupta, G. M. Dhar, Thermochim. Acta. , 443

(2006) 87.

115. T. Chiranjeevi, G. M. Kumaran, J. K. Gupta, G. M. Dhar, Pet. Sci. Technol. , 26 (2008)

690.

116. R. H. Bowker, M. C. Smith, M. L. Pease, K. M. Slenkamp, L. Kovarik, M. E. Bussell,

ACS Catal., 1 (2011) 917.

117. K. S. P. Rao,S. Khajamasthan,G. Muralidhar,K. V. R. Chary,V. S. Subrahmanyam, J.

Chem. Soc., Chem. Comm., (1988) 375.

118. K.V.R. Chary,K. S. Ramarao,G. Muralidhar,P. Kantarao, Letter to Editor, (1990).

119. Furimsky, E., Ind. Eng. Chem. Prod. Res. Dev., 22 (1983) 31.

120. K. Kreuzer ,R. Kramer, J. Catal., 167 (1997) 391.

121. M. Bartok,G. Szollosi,J. Apjok, React. Kinect. Catal. Lett., 64 (1998) 21.

122. S. K. Maity, M. S. Rana, S. K. Bej, J. Ancheyta-Juarez, G. M. Dhar, T. S. R. P. Rao,

Catal. Lett., 72 (2001) 115.

123. U. Gennari, R. Krammer, H. L. Gruber, Appl. Catal., 11 (1984) 341.

124. U. Gennari, R. Krammer, H. L. Gruber, Appl. Catal., 44 (1988) 239.

125. F. Notheisz ,M. Bartok, J. Catal., 71 (1981) 331.

126. M. Bartok ,F. Notheisz, J. Catal., 68 (1981) 209.

23

Chapter 2

Rake Mechanism for the Deoxygenation of Ethanol over a

Supported Ni2P/SiO2 Catalyst

D. Li, P. Bui, H. Zhao, S. T. Oyama, T. Dou, Z. H. Shen

Journal of Catalysis, Volume 290, June 2012, Pages 1-12

Used with permission of the Elsevier, 2013

2.1. Introduction

The utilization of biomass for the production of fuels and chemicals is currently an area of

great activity because of the recognition that fossil fuels are a finite resource and because of the

imminent threat of global warming from the release of carbon dioxide [1-6]. Among the major

products of biomass conversion is ethanol, a commodity chemical derived from the fermentation

of sugar cane or energy-rich crops such as corn [7, 8]. Although ethanol is produced in large

quantities as a substitute or supplement for fuels derived from crude oil, it is currently unlikely to

replace much less expensive chemical feedstocks derived from petroleum or natural gas.

Nevertheless, in a future with limited hydrocarbon supplies there may be a role for ethanol as a

primary feedstock [9] as recently demonstrated by the startup of a 200,000 metric ton ethylene

from ethanol plant in Brazil [10], and for this reason research in ethanol conversion is warranted.

It is also of interest to study ethanol for fundamental reasons. First, it is one of the simplest of

oxygenated compounds and its deoxygenation can provide insight into oxygen removal from

more complicated molecules. Second, it can react to produce different products, and for this

24

reason can be used as a probe to relate the reaction pathway to the properties of the catalysts, in

particular their acid-base properties versus their metallic nature. This study is thus motivated in

part to explore the catalytic chemistry of deoxygenation and in part to determine the role of

intermediates in the reaction pathway. Use is made of a traditional acid catalyst, HZSM-5, which

has been shown to be an effective agent for dehydration, and a metallic catalyst, Ni2P, which has

been demonstrated to have hydrogen transfer capabilities. The use of metallic catalysts is

important because studies have shown that biomass conversion cannot be carried out by acid

catalysts alone [1-3].

The decomposition of ethanol on solid catalysts typically occurs through competing

reactions [8, 11-13]: 1) intermolecular dehydration, which gives diethyl ether and water, 2)

intramolecular dehydration, which yields ethylene and water, 3) dehydrogenation which

produces acetaldehyde and hydrogen, and 4) total decomposition to CO, H2, CH4, C and O [7].

The acidity and basicity of the solid catalysts are two important factors that influence their

activity and selectivity. Dehydration generally occurs on acidic sites [14-16], and

dehydrogenation usually proceeds on basic sites [17, 18].

An important pathway for ethanol conversion is dehydration to ethylene. Previous studies

reported that HZSM-5 zeolite with strong Brønsted acid sites was an effective catalyst for this

transformation [19]. However, deactivation by formation of coke on its surface led to decreasing

activity and selectivity towards ethylene [20] and therefore made the process unsuitable for

industrial applications [21]. Phillips and Datta [22] investigated the production of ethylene from

hydrous ethanol over HZSM-5 under mild conditions and demonstrated that strong Brønsted acid

sites led to rapid catalyst deactivation in the initial stages through the oligomerization of ethylene

25

and the formation of carbonaceous species. However, moderating the acidity could reduce coke

formation and enhance the steady-state catalytic activity of HZSM-5.

Another important pathway for ethanol decomposition is dehydrogenation to acetaldehyde

and hydrogen which is favored on basic [23] or metallic catalysts [24]. Chang et al. reported that

copper catalysts supported on rice husk ash displayed high catalytic activity and selectivity

towards dehydrogenation [24]. Much attention has been focused on preparing bifunctional

catalysts with both acidic and basic sites in order to promote both dehydrogenation and

dehydration [25-28]. Aramendía et al. [29, 30] claimed that basic sites were responsible for

dehydrogenation and both weak acid and basic sites were associated with dehydration.

Although there are a wide variety of catalysts for ethanol conversion, there has been no

report on the decomposition of ethanol over transition metal phosphide catalysts. Transition

metal phosphide catalysts possess excellent hydrogen transfer properties and have been

investigated extensively in hydrotreating applications. They have also been used in

hydrodeoxygenation studies and have shown promising results [31]. Therefore, it is of interest to

determine the activity of transition metal phosphides in ethanol conversion. In the present study,

Ni2P/SiO2 catalyst was chosen for catalyzing the ethanol decomposition process because it is

generally the most active of the transition metal phosphides for a variety of reactions (HDS,

HDN, and HDO) [31-39]. Comparison is made to HZSM-5, the most widely studied acid

catalyst.

2.2. Materials and Experiments

26

2.2.1. Materials

The HZSM-5 (Si/Al=15) commercial catalysts were obtained from Zeolyst International.

The transition metal phosphide Ni2P/SiO2was prepared on a fumed silica EH-5 support provided

by the Cabot Corp. The chemicals used in the synthesis of the catalyst were Ni(NO3)2·6H2O

(Alfa Aesar, 99%), (NH4)2HPO4 (Aldrich, 99%). The chemical utilized in the reactivity study

was ethanol (Decon Laboratories, Inc, 200 Proof). The gases employed were H2 (Airco, Grade 5,

99.99%), He (Airco, Grade 5, 99.99%), CO (Linde Research Grade, 99.97%), 0.5% O2/He

(Airco, UHP Grade, 99.99%), O2 (Airco, UHP Grade, 99.99%), 10% CO2 (Airco, UHP Grade,

99.99%), NH3 (Alexander Chemical Corporation, AH200).

2.2.2. Nickel phosphide (Ni2P/SiO2) synthesis

The Ni2P/SiO2 catalyst was prepared by temperature-programmed reduction (TPR),

following procedures reported previously [35, 38, 39]. Briefly, the synthesis of the catalysts

involved two stages. First, a solution of the corresponding metal phosphate precursor was

prepared by dissolving appropriate amounts of Ni(NO3)2·6H2O with ammonium phosphate in

distilled water, and the solution was used to impregnate silica by the incipient wetness method.

The obtained samples were dried and calcined at 773 K for 6 h, then ground with a mortar and

pestle, pelletized with a press (Carver, Model C), and sieved to particles of 650–1180 µm

diameter (16/20 mesh). Second, temperature-programmed reduction (TPR) was carried out on

the pelletized precursor phosphate (typically 200mg) placed in a U-shaped quartz reactor. The

sample was heated from room temperature to 1073 K at 2 K min-1

in flowing hydrogen at 100

cm3

(NTP) min-1

g-1

to reduce the metal phosphate to the desired phosphide and to determine the

27

peak maximum for reduction. A portion of the exit gas flow was sampled through a leak valve

into a mass spectrometer (MS, Dycor/Ametek, model MA100) and the masses 2(H2), 18(H2O),

31(P) and 34(PH3) were monitored during the experiment. For large sample sizes sample was

kept at the determined peak reduction temperature (883 K) for 2 h, followed by cooling to room

temperature under He flow [100 cm3 (NTP) min

−1], and then passivated at room temperature in a

0.5% O2/He for 4 h. The Ni molar loading was 1.156 mmol per g of support, corresponding to a

weight loading of Ni2P of 7.9 wt% with an initial Ni/P ratio of 1/2.

2.2.3. Characterization

Irreversible CO uptake measurements were used to provide an estimate of the active sites

on the catalysts. Usually, 200 mg of a passivated Ni2P/SiO2 catalyst was loaded into a U-shaped

quartz reactor and then the sample was reduced in flowing H2 [200 cm3 (NTP) min

−1] at 723 K

for 2 h. For HZSM-5 the pretreatment temperature was 773 K using He as the purging gas. After

cooling in He [200 cm3 (NTP) min

−1], pulses of CO in a He carrier at 43 µmol s

−1 [65 cm

3 (NTP)

min−1

] were injected at room temperature through a sampling valve. The mass 28 (CO) signal

was monitored with a mass spectrometer (MS, Dycor/Ametek, model MA100). CO uptake was

calculated by measuring the decrease in the peak areas caused by adsorption in comparison with

the area of a calibrated volume (19.5 µmol).

Surface areas of the samples were obtained using the BET method based on adsorption

isotherms at liquid nitrogen temperature, using a value of 0.162 nm2 for the cross-sectional area

of a N2 molecule. The measurements were performed in a volumetric adsorption unit

(Micromeritics ASAP 2000). X-ray diffraction (XRD) patterns of the samples were obtained with

28

a PANalyticalX’pert Pro powder diffractometer operated at 45kV, using Cu Kα

monochromatized radiation (λ=0.154178nm).

Transmission infrared spectra of pyridine adsorbed on the Ni2P/SiO2 catalyst and HZSM-

5(Si/Al=15) were collected to characterize the surface acidic sites. Fourier transform infrared

(FTIR) spectra measurement were carried out with a Digilab Excalibur Series FTS 3000

spectrometer equipped with a liquid N2 cooled mercury-cadmium-telluride detector. The IR cell

was equipped with water cooled KBr windows, connections for inlet and outlet flows, and

thermocouples connected to a temperature controller to monitor and control the sample

temperature. For the experiments, about 25 mg of finely ground Ni2P/SiO2 catalyst, or HZSM-5

samples were pressed into self-supporting wafers with a diameter of 13 mm (18.8 mg cm-2

).

Wafers were mounted vertically in a quartz sample holder to keep the incident IR beam normal

to the sample. Before dosing pyridine, the sample of Ni2P/SiO2 was reduced in H2 and HZSM-5

was purged in He flow at 723 K and 773 K, respectively, for 2 h at a flow rate of 100 µmol s−1

[150 cm3

(NTP) min−1

]. After pretreatment, the samples were slowly cooled in flowing He [200

cm3

(NTP) min-1

] and background spectra were collected at 423K under He flow. The samples

were dosed at atmospheric pressure and room temperature with 1.0 mol % pyridine in He carrier

at a total flow rate of 140 µmol s-1

[200 cm3

(NTP) min-1

] until saturation was achieved. The

samples were then purged with carrier gas for 0.5 h to remove gaseous and weakly adsorbed

pyridine. Spectra were acquired at 423 K and are shown with subtraction of the background

contribution to highlight the pyridine adsorbate peaks.

Ethanol FTIR measurements were carried out with Ni2P/SiO2 catalyst in the same manner

as with the pyridine experiments. After pretreatment background spectra were collected under He

flow, and then ethanol was introduced with He flow until saturation was achieved.

29

Ammonia, pyridine and carbon dioxide temperature-programmed desorption (NH3-TPD,

pyridine-TPD and CO2-TPD) were used to determine the quantities of acid and base sites on the

catalysts. TPD experiments were carried out by loading catalyst samples (200 mg) in U-shaped

quartz reactors connected to a MS pretreating as in the CO uptake experiments, and then dosing

NH3 or CO2 or a mixture of 2.6 mol% pyridine in dry He flow from a bubbler containing pyridine

at room temperature onto the samples at room temperature until saturation. Then samples were

purged with pure He for 1 h at a flow rate of 200 cm3(NTP) min

−1 to remove physisorbed NH3 or

pyridine or CO2, then heated from room temperature to 973 K at 10 K min-1

. The masses

17(NH3) or 79(pyridine) or 44(CO2) were monitored by the MS during the experiments. Peak

areas were quantified by comparison to the areas of calibrated pulses of the pure compounds.

Ethanol temperature programmed desorption (EtOH-TPD) was used to compare the

catalytic activity and product selectivity of ethanol over the two different catalysts. To start a

reaction, catalysts (200 mg) were placed in a U-shaped reactor and pretreated at the same

conditions as used for chemisorption. After pretreatment, a flow of He saturated with ethanol

was introduced at 200 cm3

(NTP) min-1

for 1 h at room temperature. After purging in He flow for

1 h, the temperature was raised at 10 K min-1

to 1073 K in He flow [200 cm3

(NTP) min-1

]. The

reaction products were analyzed using an online mass spectrometer (Dycor/Ametek, model

MA100). The conversion of ethanol and the product selectivity for each sample were calculated

and compared. The products were quantified using reaction response factors (RRF). The RRF

were determined experimentally calibrating the mass spectrometer signals with pure standards of

known concentration.

30

2.2.4. Reactivity Studies

Activity tests were conducted in a U-shaped reactor at atmospheric pressure using catalyst

amounts corresponding to 121 µmol of active sites. For nickel phosphide the catalyst amount

(903 mg) was based on the CO uptake (134 µmol g-1

) and for HZSM 5 the catalyst amount (214

mg) was based on the NH3 uptake (565 µmol g-1

). To start, the catalyst was pretreated at the same

conditions as used for chemisorption. After pretreatment, a mixture of 1.5 mol% ethanol in dry

He flow from a bubbler containing ethanol at 273 K was passed onto the sample, and then the

catalyst was stabilized for 0.5 h after the feed was introduced. Reactivity testing was performed

as a function of temperature, starting at the highest temperature of 523 K and was varied

downwards and upwards with the initial temperature repeated at the end, to establish catalyst

stability. Mass balances closed within 5%.

Reactivity tests were also investigated as a function of contact time. The contact time is

defined by the following equation, with the quantity of sites obtained from CO uptake

experiment.

Contacttime[s] =Quantityofsites[μmolg��] × Catalystweight[g]

Reactantflowrate[μmols��]

The experiment started at the highest contact time of 270 s and was varied downwards and

upwards with the initial contact time repeated at the end. Generally it took about 10 h of on

stream time to collect the rate data at several reaction temperatures or contact times, and the

smoothness of the data indicated no deactivation. The reaction products were analyzed using an

online mass spectrometer (Dycor/Ametek, model MA100). The conversion and product yield

31

were evaluated by determining the amount of ethanol reacted and products formed using the

formulas below.

Conversion[%] =N(ethanol)"# − N(ethanol)%&'

N(ethanol)"#× 100

Yield[%] = Fractionalconversion × Selectivity × 100

2.3. Results and Discussion

2.3.1. Synthesis and basic characterization

The supported nickel phosphide was prepared in two stages as described in the

experimental part. First, solutions of the nickel and phosphorous components were impregnated

on the silica support and the material was dried to form supported phosphate precursors. Second,

the phosphate was transformed into a phosphide by temperature-programmed reduction (TPR).

The TPR experiment was carried out to understand the phenomena involved in the reduction

process and to determine the optimum reduction condition used for large scale catalyst

preparation (Fig. 2.1a). A sample of HZSM-5 was also included in the studies because it is

among the most effective of reported catalysts, and serves as a comparison to the Ni2P material.

Modification of ZSM-5, though interesting, would detract from the treated subject.

Only the results for mass 18 (H2O) are shown, because the other monitored masses

provided little additional information. The water evolution (mass 18) shows a reduction peak at

32

883 K. This temperature was chosen for preparation of larger scale samples used for

characterization and testing.

Analysis of the product of TPR was carried out by XRD (Fig. 2.1b). The diffraction pattern

for the silica supported nickel phosphide shows three major peaks at 40.5°, 44.8° and 47.5°,

which line up well with the standard pattern for Ni2P. The peaks are significantly broadened

indicating that small Ni2P crystals were formed. The XRD pattern after reaction was essentially

unchanged, indicating that the catalyst was stable.

The width at half-maximum was corrected for instrumental broadening (0.2°) and the

crystallite size of the supported Ni2P was calculated using the Scherrer equation, ./ =01

23%4(5),

where K is a constant taken as 0.9, λ is the wavelength of the X-ray radiation (0.154178 nm), β is

the peak width in radians at half-maximum, and 2θ is the Bragg angle. Table 2.1 reports the true

width at half-maximum of the peak and the crystallite size.

400 500 600 700 800 900 1000 1100

883K

Ma

ss S

pe

ctr

om

ete

r S

ign

al / A

.U.

(m=

18

)

Temperature / K

2θ / Degrees 20 30 40 50 60 70

PDF reference

Ni2P/SiO

2

Inte

nsity / a

.u.

a) Mass 18 signal from TPR b) X-ray diffraction pattern

Fig. 2.1. Synthesis and XRD Analysis of Ni2P/SiO2.

33

Table 2.1.The true width at half-maximum of the peak and

crystallite size of Ni2P/SiO2 catalyst

Peak / degree Direction True width / rad Crystallite size / nm

40.5 111 0.009 17

44.8 201 0.007 22

47.5 210 0.006 25

54.2 300 0.010 15

Average -- 0.008 20

There is no substantial anisotropy with crystallographic direction, indicating that the

crystallites are spherical, in agreement with high resolution transmission electron microscopy

results from the group of Bussell [40].

The CO chemisorption and BET characterization results are reported in Table 2.2. Earlier

studies have shown that uptakes of CO on SiO2 and Al2O3 were negligible [41-43]. The CO

chemisorption uptake of the Ni2P/SiO2 catalyst was 134 µmol g-1

, which is in line with previous

measurements. The CO uptake of HZSM-5 was small, and was probably due to CO interaction

with strong Lewis acid sites, as HZSM-5 does not contain metal. Both the phosphide and the

HZSM-5 samples have high BET surface areas, 310 and 410 m2

g-1

, respectively. The BET

surface area of Ni2P/SiO2 is lower than that of the SiO2 support of 350 m2 g

-1, and this was likely

caused by sintering during the preparation process. Additionally, Table 2.2 also lists the pore

volumes of the two catalysts obtained from N2 physisorption measurements and shows that the

major types of pores in Ni2P/SiO2 and HZSM-5 catalysts are mesopores and micropores,

respectively.

34

It has been reported [44] that micropores are associated with coke formation that results in

the zeolite's loss of activity during the ethanol dehydration process. The micropore size limits the

size of the channels, cavities, and channel intersections, trapping the species responsible for

coking. Mesopores are more active in the ethanol reaction, for the larger size allows easier

passage of the reactant for reaching the active sites and for the products to leave. Gayubo et al.

[45] studied the selectivity of olefin products from bioethanol and demonstrated that higher

mesoporosity and moderate acid strength were suitable for the decomposition of bioethanol.

2.3.2. Temperature-programmed desorption (TPD) of NH3, pyridine, and CO2

Measurements of NH3 [46, 47], pyridine-, and CO2- [48, 49] temperature - programmed

desorption (TPD) were carried out in order to clarify the relationship between the catalyst

activity and the amount of acidic and basic sites on the catalysts. The results are shown in Figs.2-

4 and summarized in Table 2.3. Although the signal is reported with arbitrary units, the peaks

were quantitated by injection of known quantities of NH3, pyridine, and CO2.

Relevant to the NH3-TPD results, previous studies showed that [50] peak maxima, Tm,

from 293 K to 473 K correspond to weak acidity sites,Tm from 473 K to 673 K to intermediate

acidity sites, and Tm higher than 673 K to strong acidity sites. The HZSM-5 sample shows two

peaks at Tm of 500 K and 724 K, respectively (Fig. 2.2), clearly suggesting the presence of

Table 2.2. Characterization results for HZSM-5 and Ni2P/SiO2 catalysts

Catalyst CO-Uptake BET area Pore Volume (cm3 g

-1)

µmolg-1

m2 g

-1 Vmicro

a Vmeso

b Vtot

Ni2P/SiO2 134 310 0.009 0.083 0.092

HZSM-5 11 410 0.124 0.036 0.160 aMicropore volume from t-plot

bVtotal-Vmicro

35

moderate and strong acidic sites. The ratio of the moderate acidic sites to the strong ones is 0.75.

Bi et al. [51] and Post and Hoof [52] reported that ethylene was easily polymerized on the strong

Brønsted acid sites of HZSM-5, and suggested that elimination of the strong acidic sites could

improve the stability of the catalyst. The Ni2P/SiO2 catalyst exhibits one broad peak from 402 K

to 805 K with a maximum around 475 K, indicating that the catalyst has plenty of moderate

acidity sites. The acid sites are probably associated with phosphorus on the support. In summary,

the NH3-TPD characterization results show that Ni2P has weak and moderate strength acid sites,

while HZSM-5 has stronger acid sites the strongest of which might lead to side reactions.

400 500 600 700 800 900 1000

Temperature / K

Ma

ss S

ep

ectr

om

ete

r S

ign

al / A

.U.

(m=

17

)

ZSM-5

Ni2P/SiO

2724K

500K

475K

Fig. 2.2. Mass 17 signal from NH3-TPD

Previous references noted that [53, 54] pyridine is a strong base in the gas phase and

requires higher temperature for complete desorption than ammonia from catalysts. Figure 3

displays the pyridine-TPD profiles and shows that all peak maxima occur at higher temperature

than the corresponding NH3-TPD maxima (530 K for Ni2P/SiO2, 514 K and 991 K for HZSM-5,

respectively). The conclusions are similar to those obtained with NH3-TPD, with Ni2P/SiO2

36

exhibiting moderate strength acidic sites and HZSM-5 possessing both moderate and strong

acidic sites in a ratio of 0.60.

Figure 2.4 shows the TPD profiles of CO2 from the Ni2P/SiO2 and HZSM-5 samples, with

peak maxima located at 389 K and 373 K, respectively. Table 2.3 summarizes the quantity of

acidic and basic sites. The results suggest that Ni2P/SiO2 has fewer acidic sites and more basic

sites than HZSM-5. Previous FTIR characterization of acid sites indicate that there are Si-OH

and P-OH groups on the support that contribute to the total acidity [55-57]. The quantity of

probe molecules on HZSM-5 is generally larger than on the Ni2P/SiO2, reflecting the large

contribution of the total surface of the material to the adsorption.

400 600 800 1000 1200

Ni2P/SiO

2

530k

ZSM-5Mass S

epectrom

ete

r S

ignal / A

. U

.

(m

=79)

Temperature / K

991K

514K

Fig. 2.3. Mass 79 signal from pyridine-TPD

37

The influence of acidic and basic centers on dehydration has been well studied. Basic and

acidic sites are necessary for the dissociative adsorption of ethanol [58, 59]. Golay et al. [60]

investigated the influence of the catalyst's acid/base properties on the catalytic ethanol

dehydration. They suggested that for the formation of ethylene a simultaneous adsorption on

both acidic sites and basic sites was necessary and that the dehydration was affected by both the

surface acidity and by the sorption kinetics on the basic sites.

2.3.3. Infrared spectroscopy of pyridine

Pyridine adsorption is often combined with in situ Fourier transform infrared spectroscopy

(FTIR) to probe the surface acidic properties of supports and catalysts [61, 62]. Characteristic

bands in the FTIR spectrum are used to determine if pyridine is protonated through the nitrogen

atom by surface Brønsted acid sites and/or bonded to coordinative unsaturated metal sites (Lewis

acids). Upon interaction with a Brønsted acid site, pyridine is protonated to a pyridinium ion and

Fig. 2.4.. Mass 44 signal from CO2-TPD

400 500 600 700 800 900 1000

Ni2P/SiO

2

ZSM-5

Mass S

pectrom

ete

r S

ingal / A

.U

(m=44)

Temperature / K

373 K

389 K

38

absorbs with a characteristic band around 1545 – 1540 cm-1

. Interaction of pyridine with Lewis

acid sites leads to a coordinatively bonded pyridinium complex with a well-resolved band

centered around 1452 – 1447 cm-1

. A band located around 1490 cm-1

is common to both

adsorbed species. Many studies have employed this technique to qualitatively and quantitatively

study the acidic properties of catalytic materials and the resulting effect on catalyst properties

and activity. Figure 2.5 displays FTIR spectra of adsorbed pyridine in He flow on HZSM-5and

Ni2P/SiO2 catalysts at 423Kand shows that both possess acidic sites. HZSM-5 contains mostly

Brønsted acid sites whereas Ni2P/SiO2 has more Lewis acid sites.

1560 1540 1520 1500 1480 1460 1440 1420 1400

Ab

so

rba

nce

/ A

.U

Wavenumbers / cm-1

ZSM-5

Ni2P/SiO

2

The quantity of Brönsted acid sites and Lewis acid sites were calculated using the

following equations [63], and the results are presented in Table 3.

C(B) = IMEC(B)�� × IA(B) × πR</W

C(L) = IMEC(L)�� × IA(L) × πR</W

Where: C: Concentration (µmol/g catalyst).

Fig. 2.5. Infrared spectra of adsorbed pyridine over samples at 423 K

39

IMEC(B, L): Integrated molar extinction coefficients (cm/µmol).

IA(B, L): Integrated absorbances (cm-1

).

R: Radius of catalyst disk (cm).

W: Weight of disk (mg)

Table 2.3. Acid and base properties for HZSM-5 and Ni2P/SiO2 catalysts

Samples CO2-TPD NH3-TPD Pyridine-FTIR

µmol g-1

µmol g-1

B / µmol g-1

L / µmol g-1

B+L / µmol g-1

B/L

Ni2P/SiO2* 408 381 36 290 326 0.12

HZSM-5 218 565 428 58 486 7.38

* The Ni2P/SiO2 CO2-TPD amount is corrected by 149 µmol g-1

from the SiO2 support and the

NH3-TPD amount is corrected by 229 µmol g-1

from the SiO2 support.

Comparing the quantity of acid sites obtained by the TPD and FTIR methods, it is found

that on Ni2P the NH3 TPD amount (381 µmol g-1

) is slightly larger than the pyridine FTIR

amount (326 µmol g-1

). This also is found on HZSM-5 where the TPD amount (565 µmol g-1

) is

larger than the pyridine FTIR amount (486 µmol g-1

). There are two possible reasons for this.

First, ammonia is smaller than pyridine, so it will pack more closely on the surface as well as

enter more readily into smaller micropores. Second, the purge temperature of ammonia was

room temperature whereas that of pyridine in the FTIR measurements was 423 K, so that some

weakly bound pyridine was desorbed.

Phillips and Datta [22] reported that Brønsted acid sites were involved in the HZSM-5

catalyzed ethanol dehydration to ethylene; however, oligomerization of ethylene also occurred at

40

the same sites forming carbonaceous deposits that covered these sites and dramatically reduced

the catalytic activity. Increasing Lewis acid sites and decreasing Brønsted acid sites were

favorable for the catalytic decomposition of ethanol [21, 64].

The number of sites titrated by CO2 is surprisingly high. For Ni2P the quantity (408 µmol

g-1

) exceeds the acidic sites (381 µmol g-1

) and the surface metal sites from CO chemisorption

(134 µmol g-1

). For HZSM-5 the quantity of CO2 adsorption (218 µmol g-1

) is about 1/3 the total

acid amount (565 µmol g-1

). The adsorption of CO2 is usually attributed to basic sites, but in this

case for Ni2P there is likely to be some adsorption on the metallic sites. For both Ni2P and

HZSM-5 there are likely contributions from physisorption.

2.3.4. Ethanol temperature-programmed desorption

Figures 2.6a and b show the signal of ethylene and acetaldehyde, respectively, obtained

from ethanol temperature programmed desorption (EtOH-TPD). Figure 2.6a shows that ethylene

formation over HZSM-5 occurred at around 419 K and 533 K, and over Ni2P/SiO2 occurred at

around 407 K and 584 K. Thus, ethylene started to form and desorb from the Ni2P/SiO2 catalyst

at a lower temperature than from HZSM-5.

Figure 2.6b reveals the desorption of acetaldehyde. The formation of acetaldehyde over

HZSM-5 occurred at around 422 K and 541 K. Over Ni2P/SiO2 acetaldehyde desorbed at 411 K

and 497 K.

41

a. Signal of ethylene from EtOH-TPD b. Signal of acetaldehyde from EtOH-TPD

Fig. 2.6. Infrared spectra of adsorbed pyridine over samples at 423 K

Table 2.4 summarizes the distribution of ethylene and acetaldehyde over different

temperatures in the EtOH-TPD experiments over HZSM-5 and Ni2P/SiO2 catalysts.

Over HZSM-5, strong ethylene and acetaldehyde desorption were recorded at 530 K and

541 K, respectively. Over Ni2P/SiO2 an almost similar amount of ethylene was observed at 407

Table 2.4. Distribution of ethylene and acetaldehyde in different temperature ranges

on EtOH-TPD

Catalyst HZSM-5 Ni2P/SiO2

Maximum

Temperature / K 419 422 533 541

407 411 497 584

Percentage

of ethylene / % 20 -- 80 --

50 -- -- 50

Percentage

of acetaldehyde / % -- 16 -- 84

-- 28 72 --

300 400 500 600 700

Ma

ss S

ep

ctr

om

ete

r S

ign

al / A

. U

.

Temperature / K

ZSM-5

Ni2P/SiO

2

419 K

407 K

533 K

584 K

300 400 500 600 700

Temperature / K

Ma

ss S

ep

ctr

om

ete

r S

igna

l / A

. U

.

422 K

411 K

541 K

497 K

42

K and 584 K, and acetaldehyde desorption occurred primarily at 497 K, with a smaller peak at

411 K. The Ni2P/SiO2 catalyst shows reactivity at lower temperature than HZSM-5 particularly

for acetaldehyde, which indicates it may be an initial product. This will be confirmed later.

Figure 2.7 shows the conversion of ethanol and the selectivity towards products from

EtOH-TPD over the two catalysts.

Fig. 2.7. Conversion and Selectivity from EtOH-TPD

Overall, Ni2P/SiO2 has a little higher conversion of ethanol (57%) than HZSM-5 (56%),

but also exhibits a higher selectivity towards acetaldehyde (21%) than HZSM-5 (4%), and a

lower selectivity towards ethylene (57%) than HZSM-5 (96%). At the same time, carbon

monoxide, a product of conversion from acetaldehyde was detected at up to 20% in the product

stream on the Ni2P/SiO2 catalyst along with a small quantity of methane (3%). In contrast, there

was no carbon monoxide and methane detected from HZSM-5. The decomposition of ethanol

0

20

40

60

80

100

0

20

40

60

80

100

Ni2P/SiO

2ZSM-5

Sele

ctiv

ity / %

Co

nve

rsio

n /

%

Conversion of ethanol

Acetaldehyde selectivity

Methane selectivity

Ethylene selectivity

Carbon monoxide selectivity

43

has been studied extensively on different metallic catalysts such as Pt/Al2O3 [65], CuO with 5%

CoO and 1% Cr2O3 [66], Pt/ZrO2 [67] and supported Au [68], and it has been reported that the

dehydrogenation of ethanol yielded acetaldehyde and hydrogen: C2H5OH → CH3CHO + H2 (1).

It is likely that the methane detected with the Ni2P/SiO2 came from an intermediate acetaldehyde

which decomposed to methane and carbon monoxide: CH3CHO → CH4 + CO (2). It should be

noted that in the presence of water (a dehydration by-product) the CH4 steam reforming reaction

may also occur to produce CO and H2: CH4 + H2O → CO + 3H2 (3). Farkas and Solymosi [69]

studied the adsorption, desorption and dissociation of ethanol on MoC/Mo(100). They obtained a

product mixture of 18% of ethylene, 12% of acetaldehyde and 44% of carbon monoxide. The

formation of methane, however, was not reported. They suggested the reason for no methane

formation was that no CHx species formed from rupture of a C-C bond on the catalyst surface

during the dehydrogenation process. These results are in agreement with the finding from this

study that the amount of methane was much smaller than that of carbon monoxide.

Under our conditions of EtOH-TPD, the Ni2P/SiO2 displayed better ability for the

conversion of ethanol and selectivity towards dehydrogenation than HZSM-5. Dehydrogenation

is generally accepted to occur on basic sites and metallic sites. On Ni2P the uptake of CO2 is

substantial (Table 3), but as discussed earlier, some of this may be due to chemisorption on

metallic sites, and a considerable portion in physisorbed mode. It is more likely, from the known

hydrogen transfer capabilities of the phosphide that dehydrogenation is occurring on metallic

sites.

In order to ensure that the TPD results were not affected by the lag time for gas to diffuse

out of pores, an analysis was carried using the criterion of Gorte [70] and Ibok and Ollis [71].

These researchers suggested that the effect of diffusion limitations could be ignored for a value

44

of less than 0.01 for the group@ABC

(DE�DF)G, where β is the heating rate (K/s), l is the width of

catalyst slab (cm), ε is the porosity (cm3/cm

3), Tf and T0 are the final and initial temperature (K),

respectively, and D the is effective diffusivity (cm2/s). For our conditions a highest value of 6×

10-5

is obtained (Table 5), and this indicates no diffusion limitations during the TPD experiments.

Table 2.5. Parameters in the Gorte Criterion

β Heating rate / Ks-1

0.17

l Width of catalyst slab / cm 0.09

ε Porosity / cm3cm

-3 0.45

Tf－T0 Temperature difference / K 110

D Effective diffusivity 0.1

2.3.5. Reactivity

The conversion of ethanol as a function of temperature was compared for HZSM-5 and

Ni2P/SiO2 with equal sites loaded in the reactor (121 µmol g-1

). For HZSM-5 the main product

at all temperatures was ethylene, with some acetaldehyde and a butylene isomer (not identified)

also formed. On Ni2P/SiO2 the main product was initially acetaldehyde, but this decreased with

increasing temperature and ethylene became the main product at high temperature. These results

are consistent with previously studies [25-30]. In general Ni2P/SiO2 exhibited higher activity

than HZSM-5 at all temperatures. Turnover frequencies were obtained from:

(Turnoverfrequency[s��] =JKL3'L#'MA%NOL'K[PQ%A4RS]×T%#UKO4"%#

V&L#'"'W%M4"'K4[PQ%AXRS]×TL'LAW4'NK"XY'[X])

45

A summary is provided in Table 2.6. Calculations of the Weisz–Prater Criterion (CWP)

were made to ascertain that no mass transfer limitations were present. The highest CWP gave

0.030 « 1, indicating that internal mass transfer effects can be neglected at the conditions

employed for the reactivity study in this paper.

Table 2.6. Reactivity in ethanol deoxygenation (Contact time = 162 s)

Catalyst Conversion / %

448 K 473 K 498 K 523 K

Ni2P/SiO2 41 54 86 100

HZSM-5 9 25 68 91

Turnover frequency / s-1

Ni2P/SiO2 0.0025 0.0033 0.0053 0.0062

HZSM-5 0.00057 0.0016 0.0042 0.0056

Apparent activation energies were calculated from fits to the conversion curves at low

conversion and were 39 kJ mol-1

for HZSM-5 and 46 kJ mol-1

for Ni2P/SiO2.

46

Fig. 2.8. Variation of ethanol conversion and product yield as a function of reaction

temperature at a contact time of 162 s. a) HZSM-5 b) Ni2P/SiO2

Figure 2.9 shows the ethanol conversion and product yields as a function of contact time

over the HZSM-5 catalyst at 473 and 498 K. It can be seen that the conversion increased with

time and temperature, as expected. The product formed in highest yield was ethylene. At 473 K

(Fig. 2.9a) the ethylene decreased slightly at high contact time, and this was accompanied by a

corresponding growth in butylenes, indicating that part of the ethylene might have been

consumed to form the C4 products. Acetaldehyde was produced at low contact time together

with ethylene, suggesting that these C2 species were produced in parallel. However,

acetaldehyde did not grow appreciably with time suggesting that it was an intermediate that also

reacted to form the butylenes. At 498 K (Fig. 2.9b) ethylene growed monotonically with contact

time. Acetaldehyde was formed in small amounts and went through a shallow maximum at short

contact time, confirming that it is an intermediate. The butylenes were formed in small but

growing amounts indicating that they were final products. Overall, the results indicated that

0

20

40

60

80

100

0

20

40

60

80

100a) HZSM-5

523K498K473K448K

}Acetaldehyde Butylene

Ethylene

Conversion

Yie

ld / %

Temperature / K

Co

nve

rsio

n /

%

448K 473K 498K 523K0

20

40

60

80

100

0

20

40

60

80

100b) Ni

2P/SiO

2

Yie

ld / %

Co

nve

rsio

n /

%

Temperature / K

Conversion

Ethylene

Acetaldehyde

47

ethylene and acetaldehyde are formed in parallel, and react by a condensation reaction to form

butylenes. At higher temperatures the ethylene was the preferred product and the acetaldehyde

was largely consumed to form the butylenes resulting in a smaller acetaldehyde yield than at

lower temperatures.

Fig. 2.9. Variation of ethanol conversion and product yield as a function of contact time on

HZSM-5. a) Reaction temperature of 473 K b) Reaction temperature of 498 K

Figure 2.10 shows the ethanol conversion and yields of acetaldehyde and ethylene as a

function of contact time at 473 and 498 K on Ni2P/SiO2. The curves are fits to the data that will

be discussed later. The figure indicates that conversion and the yields of products increase with

contact time. The data at 473 K (Fig. 2.10a) show both ethylene and acetaldehyde growing with

time. From these data alone it is not possible to deduce the sequence of reaction steps. Indeed, it

could be that acetaldehyde and ethylene are produced in parallel. As will be shown, this is not

the case. The data at 498 K (Fig. 2.10b) this time shows that acetaldehyde goes through a

maximum, while ethylene rises monotonically, behaviors characteristic of a sequential network.

This result is in accordance with a previous report [72] in which the conversion of ethanol

0 50 100 150 200 250 3000

20

40

60

80

100

0

20

40

60

80

100

Con

ve

rsio

n /

% Yie

ld / %

Contact time / s

Conversion

EthyleneAcetaldehyde Butylene

a) HZSM-5

T = 473 K

0 50 100 150 200 250 3000

20

40

60

80

100

0

20

40

60

80

100

Contact time / s

Yie

ld / %

Con

vers

ion /

%

Conversion

Ethylene

Acetaldehyde Butylene

b) HZM-5

T = 498 K

}

48

increased with increasing contact time along with a drop in acetaldehyde selectivity. As will be

discussed in the modeling section, both sets of data can be described by a reaction sequence

involving adsorbed intermediates undergoing consecutive reactions.

It should be emphasized that the measurements were carried out by varying the contact

time up and down over a period of 10 h so that the smoothness of the data provide evidence for

the lack of deactivation. This fact and the establishment of a mass balance of 100 % ± 5%

indicate that the maximum in acetaldehyde yield is not due to its consumption to form products

such as butylenes (not detected) or heavy materials like polymers.

It should be noted that no ethane was detected. At the conditions of this study the

concentration of the ethanol reactant is low (1.5 mol%), and since the concentration of products

is significantly lower, it is difficult for the hydrogen (from dehydrogenation) and ethylene (from

dehydration) to come into contact to form ethane. Furthermore, the ethane formation reaction is

favored by high temperature and pressure, which are not the conditions employed in our study.

Thus, the lack of formation of ethane is understandable. Other studies report acetaldehyde and

ethylene formation but not ethane [20, 26].

49

a) Reaction temperature of 473 K b) Reaction temperature of 498 K

Fig. 2.10. Variation of ethanol conversion and product yield as a function of contact

time on Ni2P/SiO2. The curves are fits from a simulation model discussed in the text.

Reaction temperature and contact time play important roles in the decomposition of

ethanol, and their effect on the steady-state reactivity of HZSM-5 and Ni2P/SiO2 were examined.

To summarize, on HZSM-5, the main product was ethylene with small amounts of acetaldehyde

and butylenes. Butylenes formed at long contact time and could be attributed to condensation of

ethylene and acetaldehyde on strong Brønsted acid sites on the surface of HZSM-5. On

Ni2P/SiO2 the main products of the ethanol reaction were ethylene and acetaldehyde with small

amounts of carbon monoxide and trace quantities of methane, without ethane. Acetaldehyde was

formed as an intermediate compound.

The results in Fig. 2.10 suggest that there is a consecutive reaction: ethanol �

acetaldehyde � ethylene. This seems odd, because this would entail the oxidation of ethanol

before the formation of ethylene, which is unlikely at reductive conditions as used here. It was

0 50 100 150 200 250 3000

20

40

60

80

100

0

20

40

60

80

100

Acetaldehyde

Ethylene

Conversion

Yie

ld / %

Co

nvers

ion

/ %

Residence time / s

Experimental Data

Simulation Data

b) Ni2P/SiO

2

T = 498 K

0 50 100 150 200 250 3000

20

40

60

80

100

0

20

40

60

80

100

Conversion

Experimental Data

Simulation Data

Acetaldehyde

Ethylene

Yie

ld / %

Convers

ion / %

Contact time / s

a) Ni2P/SiO

2

T = 473 K

50

reasoned that a more likely possibility was that there existed a common species at the surface

that could react to form both acetaldehyde and ethylene in a sequential manner. This led to the

following scheme.

Ethanol first adsorbs onto a vacant site to form an ethoxide species. This undergoes

dehydrogenation through an alpha-H abstraction from the carbon proximal to the oxygen to form

adsorbed acetaldehyde. The adsorbed acetaldehyde can desorb or can react further through an

enolization reaction involving a beta-H abstraction from the carbon distal to the oxygen. This can

then undergo hydrogenolysis with adsorbed hydrogen on active sites to form ethylene and an

adsorbed OH group (not shown).

Scheme 2.1. “Rake” mechanism for ethanol reaction

The catalytic reaction network was simulated based on the disappearance rate of ethanol

and the formation rate of acetylene and ethylene. Ethanol, acetylene, and ethylene were noted as

A, B, and C respectively and the corresponding adsorbed ethoxy, acetaldehyde, and vinyl alcohol

are denoted as A*, B* and C*.

51

The kinetic rate law equations are the following:

1/ Z

[

\[]]

\^= −k][A][∗] + k�][A

∗]

2/ Z

[

\�b

\^= −kb�B �∗ + k�b�B∗

3/ Z

[

\�T

\^= −kT�C �∗ + k�T�C∗

The rate of consumption of A,[ ]τd

Ad

S

VrA =− has units of mol m

-2s

-1, in which V (m

3) is the

bed volume and S (m2) is the catalyst surface area. This rate is generally expressed as

)(][ YkfLrA =− where [L] (mol m-2

) is the total concentration of sites, k is a rate constant, and

f(Y) is a dimensionless function of concentrations.

The bed volume is given by V = w/ρA, where ρA (0.3 g cm-3

) is the apparent density. The

total surface area of Ni2P (m2) given by S = awf, where a (42 m

2g

-1) is the Ni2P effective surface

area, obtained from the equation CD

aρ

6= , w (0.900 g) is the weight of catalyst, f(0.079) is the

fractional weight loading of the Ni2P, ρ (7.09 g cm-3

) is the true density, and DC (20 nm) is the

average crystallite size. Finally, τ is the contact time calculated from the volume of the catalyst

bed divided by the volumetric flow rateV

v. Simplification gives

f

D

S

V

A

C

ρ

ρ

6= . The concentrations

of the intermediates as well as the vacant active sites were obtained by solving the following set

of equations, assuming that at steady state, the net rate of the formation of the intermediates is

zero.

52

4/ \[]∗]

\^= k][A][∗] − (k�] + k�)[A

∗] + k��[B∗] = 0

5/ \[b∗]

\^= kb[B][∗] + k�[A

∗] − (k�b + k�� + k<)[B∗] + k�<[C

∗] = 0

6/ \[T∗]

\^= k<[B

∗] − (k�T + k�<)[C∗] + kT[C][∗] = 0

Use was made of an active site balance,where [L] is the total concentration of active sites:

7/ [d] = [∗] + [e∗] + [f∗] + [g∗]

This kind of scheme is known in the literature as a “rake” mechanism because the

adsorption and desorption steps when written out as in Scheme 1 resemble the prongs of the

device used to gather leaves. The rake mechanism was first described by Cormerais et al. [73]

and has been suggested to apply to many types of reactions [74, 75], but in the early days of its

introduction 30 years ago the solution of the equations was not readily possible. Even to this day

many kinetic studies still use simple first-order analysis of networks that ignore the existence of

the adsorption steps. In this case the equations were solved numerically using Polymath®

software (Table 2.7) to obtain plots of acetaldehyde and ethylene yields and conversion versus

time.

Table 2.7. Rate constants used in the simulation

Temperature kA k-A kB k-B kC k-C k1 k-1 k2 k-2

K cm3 mol

-1 s

-1 s

-1 cm

3 mol

-1 s

-1 s

-1 cm

3 mol

-1

s-1

s-1

s-1

s-1

s-1

s-1

473 2.80×107

8.5 3.5×107

6.75 2.75×107

2.25 9.0 4.5 3.75 3.0

498 3.5×107

9.0 4.0×107

12.0 3.25×107 30.0 17.0 5.0 5.5 3.5

53

As can be seen by the curves shown in Figs. 2.10a and 2.10b the simulation results are in

good agreement with the experimental points, even though the behavior of the species is

different. At 473 K the acetaldehyde yield grows together with that of ethylene, while at 498 K

there is a clear maximum in the acetaldehyde yield. The rate constants have smaller values at

lower temperature as expected. The closeness of the fits indicates that the proposed mechanism

describes the physical situation accurately.

In order to further check the reasonableness of the mechanism measurements of adsorbed

surface intermediates were undertaken. Data were taken at different temperatures with ethanol

adsorbed using He as carrier (Fig. 2.11).

Fig. 2.11. EtOH FTIR over Ni2P/SiO2 with 0.15% EtOH in He flow

Table 2.8 summarizes the assignments in the spectra. At high wavenumber two sharp

negative bands at 3746 and 3665 cm-1

are due SiO-H and PO-H stretching vibrations and appear

negative because the hydroxyls are being consumed [55]. At 2990, 2940, and 2900 cm-1

a cluster

of peaks appear which are due to C-H stretching vibrations for methyl and methylene groups.

4500 4000 3500 3000 2500 2000 1500 1000

523K

498K

473K

448K

Wavenumber / cm-1

8901070

12601366

1400

15801830

1930

29002940

2990

3665

3746

2350

1750

54

Next in the center of the spectrum appear bands at 2365, 2330 due to gas-phase CO2 interference.

For the highest temperature data weak broad bands are present at 1930 and 1830 cm-1

, which are

attributed to combination bands. A small feature at 1750 cm-1

is due to adsorbed aldehyde CHO.

At 1580 cm-1

a small peak is observed which grows with temperature. This position is

characteristic of a C=C double bond. An oddly shaped feature at 1400 cm-1

followed by a

negative peak at 1366 cm-1

is due to background subtraction. A feature at 1260 cm-1

corresponds

to a CH2 wag. Peaks at 1070 and 890 cm-1

are assigned to the SiO2 support. Peaks at lower

wavenumber are obscured by interference from the silica.

Overall, in He atmosphere, the diminution of OH bands and the growth of C-H stretching

vibrations are consistent with the formation of ethoxide species bound to Si and P functionalities.

There is also evidence for an adsorbed acetaldehyde species with a band at 1750 cm-1

and the

formation of an ethylene or vinyl alcohol species contributing to the C=C signal at 1580 cm-1

.

In order to obtain more information on the behavior of the adsorbed species, a transient

experiment was carried out. The objective was to simulate the contact time measurements

described in Fig. 2.10b. The spectra of ethanol FTIR at 497 K as a function of time are shown in

Fig. 2.12.

55

4500 4000 3500 3000 2500 2000 1500 1000

300 s

60 s

0 s

15801750

Wavenumber / cm-1

Fig. 2.12. EtOH FTIR over Ni2P/SiO2 in He at 497 K

The assignments in Fig. 2.12 are the same as in Fig. 2.11. In Fig. 2.12 vertical lines have

been placed at 1750 cm-1

(position for an aldehyde functionality, HC=O) and at 1580 cm-1

(position for an olefin double bond). Examination shows that the peak at 1750 cm-1

increased

significantly at first and then disappeared with contact time. At the same time the peak at 1580

cm-1

due to an olefin (C=C) increased with time. These results are consistent with the proposed

rake mechanism in which the adsorbed acetaldehyde is an intermediate that is converted to an

ethylene precursor on the surface of the catalyst. There was no growth in the band at 2075 cm-1

due to adsorbed carbon monoxide (CO). This further indicates that the adsorbed aldehyde did

not decompose to CH4 and CO but was converted to ethylene.

The picture that emerges from the FTIR measurements is consistent with the reaction scheme

presented earlier. Ethanol adsorbs on hydroxyl species on the silica to form ethoxide species.

Substantial work [76, 77] indicates that alkoxide species on silica are highly mobile so these can

readily move to the Ni2P surface, which is excellent in hydrogen transfer reactions [37-46]. A

56

first reaction that occurs is dehydrogenation to form an adsorbed acetaldehyde (growth in band

of 1750 cm-1

). This may be assisted by basic sites present on the Ni2P surface as indicated by the

CO2 adsorption measurements (Table 2.3). Subsequently, in He the acetaldehyde undergoes an

enolization to vinyl alcohol (growth in band at 1580 cm-1

). This set of steps accounts for the

main species formed and the main catalytic pathway in the reaction. There is also a side

reaction, which is not major in the steady-state, but which proceeds in temperature-programmed

reaction, namely the formation of adsorbed CO. This is likely the result of the decomposition of

adsorbed acetaldehyde in the presence of H2 (not He) to a methyl species and CO. The hydrogen

is likely adsorbed atomic hydrogen, and the driving force is the formation of methane and

strongly bonded CO. Although not a primary path in the steady-state reaction, this pathway

accounts for the observation of CH4 and CO in the ethanol TPD spectra (Fig. 2.7).

Table 2.8. Summary of FTIR assignments

Band (cm-1

) Assignment Species Reference

3746 (neg) SiO-H stretch Si-OH [55]

3665 (neg) PO-H stretch P-OH [55]

2990, 2940,

2900

C-H stretch CH3CH3, CH2=CH-

OH CH3CHO,

CH3CH2OH,

CH2=CH2,

[78, 79]

2075 C=O CO [40, 55]

1930, 1830 Combination

1750 HC=O HCHO [76, 77]

1580 C=C CH2=CH-OH [78]

1400 OH bend [76]

1366 (neg) Background subtraction

~1310, 1260 CH2 wag CH3CH2OH [78]

~1160 C-C stretch

combination

CH3CHO,

CH3CH2OH

CH2=CH-OH

[78]

~1070 Si-O SiO2

~890 Si-O-Si SiO2

57

~990 C=C wag

C-C stretch

CH2=CH2

CH3CH3,

CH3CH2OH

[78, 79]

790-830 CH3 rock

CH2 rock

CH3CH3, CH3CHO

CH2=CH2

[76, 78]

~760 C-H bend CH3CHO [78]

~600 CH2=C out of plane CH2=CH-OH [77]

~500 C-C-O deform CH3CHO [78]

300-400 torsion CH2=CH-OH [77]

The reaction of ethanol on HZSM-5 and Ni2P/SiO2 provides information on the reaction

pathways on a typical acid catalyst and a metal catalyst. The acid catalyst, HZSM-5 carries out

simple dehydration to form ethylene. The metal catalyst also forms mainly ethylene, but through

a non-direct pathway in which an adsorbed ethoxide species is first dehydrogenated to a surface

acetaldehyde species, which undergoes enolization to a vinyl alkoxide and subsequent

hydrodeoxygenation. This route is non-direct and occurs because of the strongly metallic

hydrogenation/dehydrogenation properties of the Ni2P. It is important because the pathway may

also be involved in the reactions of more complex molecules and for this reason its participation

should be considered in the study of their conversion.

2.4. Conclusions

The experiments described in this work lead us to the following conclusions:

1. Comparison between the properties of commercial HZSM-5(Si/Al=15) and synthesized

Ni2P/SiO2 catalysts showed that HZSM-5 possesses higher surface area and stronger acidic sites

than the Ni2P/SiO2 catalyst. The Ni2P/SiO2 catalyst essentially has more mesopores and

moderate acid and basic sites. Infrared spectroscopy of pyridine reveals that acidity of HZSM-5

58

and Ni2P/SiO2 catalysts consist mainly of Brønsted and Lewis acid sites, respectively. In the

ethanol temperature programmed desorption process (EtOH-TPD), the Ni2P/SiO2 catalyst

exhibited a higher selectivity towards dehydrogenation products (acetaldehyde) than HZSM-5.

2. Steady-state reaction results confirm that the Ni2P/SiO2 produces both acetaldehyde and

ethylene, and contact time studies indicate that acetaldehyde is a primary product and ethylene is

a secondary product. The reaction, however, does not consist of an oxidation reaction followed

by a reduction reaction, but involves sequential reactions on the surface, namely, a rake

mechanism. Analysis of the reaction sequence and simulation of the results give support for this

interpretation. Measurements by in situ Fourier transform infrared spectroscopy also give

evidence for the presence of the suggested intermediate, adsorbed acetaldehyde, on the surface of

the catalyst at reaction conditions. Measurements at increasing contact times show that it

increases and then disappears, in agreement with the corresponding reactivity results.

59

References

1. A. Wawrzetz,B. Peng,A. Hrabar,A. Jentys,A. A. Lemonidou,J. A. Lercher, J. Catal., 269

(2010) 411.

2. J. Zakzeski, P.C.A. Bruijnincx, A.L. Jongerius, B.M. Weckhuysen, Chem. Rev., 110

(2010) 3552.

3. O. Casanova,S. Iborra,A. Corma, J. Catal., 275 (2010) 236.

4. T. V. Choudhary, C. B. Phillips, Appl. Catal. A: Gen., 397 (2011) 1.

5. J. J. Bozell ,G. R. Petersen, Green Chem. , 12 (2010) 539.

6. J. C. Serrano-Ruiz,D. Wang,J. A. Dumesic, Green Chem., 12 (2010) 574.

7. A. Birot,F. Epron,C. Descorme,D. Duprez, Appl. Catal. B: Environ., 79 (2008) 17.

8. J. Rass-Hansen,R. Johansson,M. Moller,C. H. Christensen, Int. J. Hydrogen Energy 33

(2008) 4547.

9. E. B. Pereira,P. R. de la Piscina,N. Homs, Biores. Tech., 102 (2011) 3419.

10. Tullo, A. H., Chem. Eng. News, 89 (2011) 24.

11. R. L. Guenard, L. C. F. Torres, B. Kim, S. S. Perry, P. Frantz, S. V. Didziulis, Surf. Sci.,

515 (2002) 103.

12. P. A. Clayborne,T. C. Nelson,T. C. De Vore, Appl. Catal. A: Gen., 257 (2004) 225.

13. B. M. Nagaraja,A. H. Padmasri,P. Seetharamulu,K. H. P. Reddy,B. D. Raju,K. S. R. Rao,

J. Mol. Catal. A, 278 (2007) 29.

14. E. A. El-Katatny,S. A. Halawy,M. A. Mohamed,M. I. Zaki, Appl. Catal. A: Gen., (1999)

83.

15. N. R. C. F. Machado,V. Calsavara,N. G. C. Astrath,C. K. Matsuda,A. P. Junior,M. L.

Baesso, Fuel, 84 (2005) 2064.

16. V. Calsavara, M. L. Baesso, N. R. C. F. Machado, Fuel, 87 (2008) 1628.

17. N. Takezawa,C Hanamaki,Kobayashi H, J. Catal., 38 (1975) 101.

18. Y. Matsumura,K. Hashimoto,S. Yoshida, J. Catal., 117 (1989) 135.

19. I. Takahara,M. Saito,M. Inaba,K. Murata, Catal. Lett., 105 (2005) 249.

20. X. Zhang,R. Wang,X. Yang,F. Zhang, Microp. Mesop. Mater. , 116 (2008) 210.

21. J. Schulz ,F. Bandermann, Chem. Eng. Technol. , 17 (1994) 179.

22. C. B. Phillips ,R. Datta, Ind. Eng. Chem. Res., 36 (1997) 4466.

23. J. M. Vohs ,M. A. Barteau, Surf. Sci., 211 (1989) 590.

24. F. W. Chang,W. Y. Kuo,K. C. Lee, Appl. Catal. A: Gen., 246 (2003) 253.

25. M. A. Aramendía,V. Borau,C. Jiménez,J. M. Marinas,A. Porras,F. J. Urbano, J. Catal.,

161 (1996) 829.

26. Y. Shinohara,T. Nakajima,S. Suzuki, J. Mol. Struct., 460 (1999) 231.

27. M. M. Doheim ,H. G. El-Shobaky, Surf. Sci. A, 204 (2002) 169.

28. F. S. Ramos,A. M. Duarte de farias,L. E. P. Borges,J. L. Monteiro,M. A. Fraga,E. F.

Sousa-Aguiar,L. G. Appel, Catal. Today, 101 (2005) 39.

29. M. A. Aramendía,V. Boráu,C. Jiménez,J. M. Marinas,A. Porras,F. J. Urbano, React.

Kinect. Catal. Lett., 65 (1998) 25.

30. M. A. Aramendíra,V. Boráu,I. M. García,C. Jiménez,A. Marinas,J. M. Marinas,F. J.

Urbano A. Porras, Appl. Catal. A: Gen., 184 (1999) 115.

31. H. Y. Zhao, D. Li, P. Bui, S. T. Oyama, Appl. Catal. A: Gen., 391 (2011) 305.

32. Y. Shu,Y. -K Lee,S. T. Oyama, J. Catal., 236 (2005) 112.

60

33. T. Kawai, K. K. Bamdo, Y. -K. Lee, S. T. Oyama, W. -J. Chun, K. Asakura, J. Catal., 241

(2006) 20.

34. X. Wang, P. Clark, S. T. Oyama, J. Catal., 208 (2002) 321.

35. S. T. Oyama, X. Wang, Y. -K. Lee, K. Bando, F. G. Requejo, J. Catal., 210 (2002) 207.

36. S. T. Oyama, T. Gott, H. Zhao, Y. -K. Lee, Catal. Today, 143 (2009) 94.

37. Oyama, S. T., J. Catal., 216 (2003) 343.

38. S. T. Oyama, X. Wang, F. Requejo, T. Sato, Y. Yoshimura, J. Catal., 209 (2002) 1.

39. S. T. Oyama, X. Wang, Y. -K. Lee, W. -J. Chun, J. Catal., 221 (2004) 263.

40. S. J. Sawhill, K. A. Layman, D. R. Van Wyk, M. H. Engelhard, C. Wang, M. E. Bussell, J.

Catal., 231 (2005) 300.

41. P.A. Clark,X. Wang,P. Deck, J. Catal., 210 (2002) 116.

42. P. A. Clark ,S. T. Oyama, J. Catal., 218 (2003) 78.

43. S.T. Oyama, Y.-K. Lee, J. Catal., 258 (2008) 393.

44. M. Guisnet ,P. Magenoux, Stud. Surf. Sci. Catal. , 88 (1994) 53.

45. A. G. Gayubo,A. Alonso,B. Valle,A. T. Aguayo,J. Bilbao, Appl. Catal. B: Environ., 97

(2010) 299.

46. P. Berteau ,B. Delmon, Appl. Catal., 70 (1991) 307.

47. W. Wang,S. P. Wang,X. B. Ma,J. L. Gong, Catal. Today, 148 (2009) 323.

48. S. R. Jagtap, Y. P. Patil, Appl. Catal. A: Gen., 341 (2008) 133.

49. A. S. Ndou,N. Plint,N. J. Coville, Appl. Catal. A: Gen., 251 (2003) 337.

50. P. Berteau ,B. Delmon, Catal. Today, 5 (1989) 121.

51. J. D. Bi,X. W Guo,M. Liu,X. S. Wang, Catal. Today, 149 (2010) 143.

52. J. G. Post ,H. G. V. Hoof, Zeolites, 4 (1984) 9.

53. D. J. Parrillo,C. Lee,R. J. Gorte, Appl. Catal. A: Gen., 110 (1994) 67.

54. H. G. Karge,V. Dondur,J. Weitkamp, J. Phys. Chem., 95 (1991) 283.

55. Y. -K. Lee ,S. T. Oyama, J. Catal., 239 (2006) 376.

56. T. Gott, S. T. Oyama, J. Catal., 263 (2009) 359.

57. S. T. Oyama, T. Gott, K. Asakura, S. Takakusagi, K. Miyazaki, Y. Koike, K. K. Bando, J.

Catal., 268 (2009) 209.

58. K. M. A. El-Salaam ,E. A. Hassan, Surf. Technol. , 16 (1982) 121.

59. F. F. Roca,L. D. Mourgues,Y. Trambouze, J. Catal., 14 (1969) 107.

60. S. Golay, L. Kiwi-Minsker, R. Doepper, A. Renken, Chem. Eng. Sci., 54 (1999) 3593.

61. R. L. Mao, T. S. Le, M. Fairbairn, A. Muntasar, S. Xiao, G. Denes, Appl. Catal. A: Gen,

185 (1999) 41.

62. M. Lenarda,M. D. Ros,M. Casagrande,L. Storaro,R. Ganzerla, Inorg. Chim. Acta., 349

(2003) 195.

63. Emeis, C. A., J. Catal., 141 (1993) 347.

64. G. A. M. Hussein ,N. Sheppard, J. Chem. Soc., Fraday Trans., 87 (1991) 2661.

65. M. Dömöka,M. Tótha,J. Raskób,A. Erdőhelyi, Appl. Catal. B: Environ., 69 (2007) 262.

66. J. Franckaerts ,G. F. Froment, Chem. Eng. Sci., 19 (1964) 807.

67. T. Yamazaki, N. Kikuchi, M. Katoh, T. Hirose, H. Saito, T. Yoshikawa, M. Wada, Appl.

Catal. B: Environ., 99 (2010) 81.

68. A. Gazsi,A. Koos,T. Bansagi,F. Solymosi, Catal. Today, 160 (2011) 70.

69. Solymosi, A. P. Farkas and F., Surf. Sci., 601 (2007) 193.

70. J.Gorte, R., J. Catal., 75 (1982) 164.

71. E. E.Ibok ,D. F.Ollis, J. Catal., 66 (1980) 391.

61

72. B. M. Abu-Zied ,A. M. El-Awad, J. Mol. Catal. A, 176 (2001) 227.

73. F.X. Cormerais,G. Perot,F. Chevalier,M. Guisnet, J. Chem. Res., S (1980) 362.

74. S. Lars, T. Andersson, J. Catal., 98 (1986) 138.

75. Stöcker, M., Mesop. Micropor. Mat. , 29 (1999) 3.

76. M. Seman,J. N. Kondo,K. Domen,R. Radhakrishnan,S. T. Oyama, J. Phys. Chem. B, 106

(2002) 12965.

77. M. Seman,J. N. Kondo,K. Domen,R. Radhakrishnan,S. T. Oyama, Chem. Lett., 11 (2002)

1082.

78. Shimanouchi, T. Molecular vibrational frequencies. NIST Chemistry WebBook, NIST

Standard Reference Database Number 69 [cited 2011 June 23].

79. M. Hawkins ,L. Andrews, J. Am. Chem. Soc. , 105 (1983) 2523.

62

Chapter 3

Synthesis of Transition Metal Phosphides via Phosphate and

Phosphite Method and their Activity in the Hydrodeoxygenation

of 2-Methyltetrahydrofuran

P. Bui, J. A. Cecilia, S. T. Oyama, A. Takagaki, A. Infantes-Molina, H. Zhao, D. Li, E.

Rodríguez-Castellón, A. J. López

Journal of Catalysis, Volume 294, October 2012, Pages 184-198

Used with permission of the Elsevier, 2013

3.1. Introduction

The need to meet stringent environmental regulations and the depletion of fossil fuel

reserves have led to interest in biomass as a renewable energy source. The utilization of biomass-

derived fuels has several advantages over the use of conventional fuels, such as reduction in

greenhouse gas emissions, localized production, and favorable economics of utilization [1-4].

One method of biomass conversion is pyrolysis – a thermal decomposition of biomass in the

absence of oxygen that yields an energy-rich liquid mixture of oxygenated aromatic and aliphatic

compounds called bio-oil [5, 6]. Bio-oil, however, has a high oxygen content (35-50 wt%) with

15-30 wt% of water that results in low heating value, immiscibility with hydrocarbon fuels, high

acidity, and chemical and thermal instability compared with hydrocarbon fuels [7, 8]. To upgrade

the bio-oil into a usable fuel it is necessary to remove the oxygen [9]. A first objective of this

work is to study the deoxygenation of a model compound, 2-methyltetrahydrofuran. This is a

63

five-atom saturated ring compound and was chosen as a model substrate because previous work

has shown that ring compounds are difficult to deoxygenate [10]. Furans are also common

products in the thermal degradation of biomass, such as pyrolysis [11]. The presence of the

methyl group allows distinguishing between the ring-opening products and hence gives

mechanistic information.

Standard catalysts for the elimination of S and N heteroatoms from petroleum feedstocks

are sulfided CoMo and NiMo; these may also be used for the removal of O, for example from

phenols and furans in biomass-derived liquids [12-14], however, the results have been modest. It

is clear that new catalysts are needed. Among new compositions that have been explored in

hydrotreating processes, phosphides of transition metals are considered potential substitutes for

the CoMo and NiMo sulfided materials [15-17]. In the past decade a plethora of works has

shown that MoP [18, 19], WP [18, 20], CoP [21, 22], Fe2P [22], and Ni2P [22-25] are highly

active for hydrodesulfurization and hydrodenitrogenation of petroleum feedstocks. Recently

their good hydrogen transfer properties have been applied to the hydrodeoxygenation (HDO) of

biomass-derived feedstocks [26-28].

The most common method for the preparation of phosphides is temperature-programmed

reduction (TPR), a simple process that can be carried out at relatively moderate temperatures

with supported or unsupported precursors [15]. With regard to the use of supports, previous

research has shown that less acidic supports such as silica or hexagonal mesoporous silica (SBA-

15 [29] and MCM-41 [24]) favor the formation of metal phosphides due to the low interaction

between the support and the precursor, usually a metal phosphate. Recently, precursors other

than phosphates have also been investigated, for example, phosphosulfides [30]. A substantial

recent development is the use of metal phosphites (e.g. Ni(HPO3H)2) instead of phosphates (e.g.

64

NiHPO4) as precursors [24]. The former have phosphorous in a lower oxidation state which

make it potentially reducible at a lower temperature; for example, using a batch reduction

method, an active Ni2P was formed at 300 oC compared to 450

oC using the traditional phosphate

method [24, 25, 31, 32].

A second objective of this work besides studying the HDO reaction is to compare the

traditional phosphate method and the new phosphite method for the preparation of a variety of

transition metals. This has not been done before. For this purpose five pairs of metal phosphides

(Ni2P, CoP, FeP, MoP, and WP) supported on silica were chosen. Their activity for HDO of the

model compound 2-methyltetrahydrofuran was evaluated and compared. Furthermore, for the

most active phosphide compounds a study of reaction pathway was carried out.

3.2. Materials and Experiments

3.2.1. Materials and catalysts

The transition metal phosphides Ni2P, CoP, WP, MoP, and FeP supported on a high

surface area fumed silica support (Cab-osil ® EH5, surface area of 334 m2

g-1

and pore volume

of 0.6 cm3

g-1

) provided by Cabot Corp. were synthesized. Sources of iron, tungsten, and

molybdenum were iron(III) nitrate (Fe(NO3)3.9H2O (Aldrich 98+%)), ammonium metatungstate

hydrate ((NH4)6W12O39.H2O (Aldrich, 99.99%)), and ammonium heptamolybdate tetrahydrate

((NH4)6Mo7O24.4H2O (Alfa Aesar, 99.98%)). For the phosphate method, sources of cobalt and

nickel were cobalt(II) nitrate (Co(NO3)2.6H2O (Aldrich, 99.99%)) and nickel(II) nitrate

(Ni(NO3)2.6H2O (Alfa Aesar, 99.6%)). For the phosphite method, the precursors of cobalt and

65

nickel were cobalt(II) hydroxide (Co(OH)2 (Aldrich 95%)) and nickel(II) hydroxide (Ni(OH)2

(Aldrich, 98+%)). The sources of phosphorus were ammonium hydrogen phosphate

((NH4)2HPO4 (Aldrich 99%)) and phosphorous acid (H2PO3H (Aldrich 99%)). The 5%

Pd/Al2O3 commercial catalyst was provided by BASF Catalysts, Inc. The model compound used

in reactivity study was 2-methyltetrahydrofuran (Aldrich 99.95%). The gases employed were H2

(Airgas, Grade 5), He (Airgas, Grade 5), CO (Linde Research Grade, 99.97%), 0.5% O2/He

(Airgas, UHP Grade), O2 (Airgas, UHP Grade), N2 (Airgas, Grade 5).

The synthesis of the supported phosphide catalysts was carried out by a standard phosphate

method and a more recently developed phosphite method. A silica support (Cab-osil ® EH5)

was employed in both methods. The support was dried at 120 oC for 6 h and calcined at 500

oC

for 4 h prior to use. Both synthetic routes required the preparation of a precursor in three steps.

First, a mixed solution of metal and phosphorus compounds was made by adding an appropriate

amount of a desired metal salt into a phosphorus solution; the phosphorous solution was made by

mixing ammonium phosphate (phosphate method) or phosphorous acid (phosphite method) into

distilled water. Second, the mixed solution was used to impregnate the silica support to the

incipient wetness point (about 2 cm3 of solution per gram of support). Finally, the obtained

phosphate precursor mixture was dried at 120 o

C for 6 h and calcined at 500 oC for 6 h; whereas

the phosphite precursor mixture was only dried in air at 80 oC overnight without any calcination.

Then all samples were pelletized and sieved to 16/20 mesh size. In the same manner, the

precursors from both methods were reduced to phosphides by a temperature-programmed

reduction (TPR) at a heating rate of 2 oC min

-1 in flowing H2 [1000 cm

3 (NTP) min

-1 g

-1] and

kept at the reduction temperature for 2 h. The reduction temperatures were determined in

separate temperature reduction experiments as the temperature at which water production peaked

66

(details are described in the characterization section). The resulting phosphides were cooled to

room temperature in He [100 cm3 (NTP) min

-1], followed by passivation in a He stream

containing 0.5% O2 for 2 h.

Table 3.1 summarizes the loadings of metal and phosphorous used in the preparation of the

catalysts. For all metals Fe, Co, Ni, Mo, and W equal moles of the element per gram of support

(1.16 mmol/g support) were used.

Table 3.1. Materials and amounts used in the preparation of phosphides supported on silica

Conventional (Phosphate) Method (A) New (Phosphite) Method (I)

Sample Metal source

(mmol/g support)

Phosphorous

(NH4)2HPO4

(mmol/g

support)

Metal

(mmol/g support)

Phosphorous

H3PO3

(mmol/g

support)

Metal

phosphide

(wt%)

Ni2P Ni(NO3)2.6H2O 1.16 2.31 Ni(OH)2 1.16 2.31 7.9

CoP Co(NO3)2.6H2O 1.16 2.31 Co(OH)2 1.16 2.31 9.4

WP (NH4)6W12O39.H2O 1.16 1.16 (NH4)6W12O39.H2O 1.16 2.31 19.9

MoP (NH4)6Mo7O24.4H2O 1.16 1.16 (NH4)6Mo7O24.4H2O 1.16 2.31 12.8

FeP Fe(NO3)3.9H2O 1.16 2.31 Fe(NO3)3.9H2O 1.16 2.31 9.1

The synthesis of the materials was carried out by temperature-programmed reduction and

monitored by mass spectrometry. The samples were characterized with carbon monoxide (CO)

chemisorption, Brunauer Emmett Teller (BET) surface area measurements, X-ray diffraction

(XRD), and X-ray photoelectron spectroscopy (XPS).

The temperature-programmed reduction was carried out in a U-shaped quartz reactor

placed in a furnace controlled by a temperature programmer (Omega Model CN 2000). The

temperature was measured with a local type K thermocouple placed near the center of the reactor

bed. An amount of 0.3 g of precursor was reduced in H2 [300 cm3 (NTP) min

-1] at a heating rate

of 2 oC min

-1 from room temperature to 800

oC. A portion of the exhaust gas flow was

67

introduced through a leak valve into a mass spectrometer (Ametek/Dycor MA100), which

monitored masses 2(H2), 4(He), 18(H2O), and 34(PH3) during the experiment. The reduction

profiles revealed reduction stages in the transformation of the precursors to the phosphides.

Irreversible CO uptake measurements were used to titrate the surface metal atoms and to

estimate the number of active sites on the catalysts. Uptakes were obtained after 0.3 g of

passivated samples were re-reduced in H2 at 450 oC for 2 h and cooled to room temperature in

He. Pulses of known amounts of CO carried by He at 30 µmol s−1

were injected into the sample.

An on-line mass spectrometer monitored the signal of mass 28 (CO) throughout the experiment.

CO uptake was calculated from the decrease in peak areas caused by adsorption of CO onto the

catalysts surface. Measurements from CO chemisorption allowed estimation of the dispersion of

metal sites using the following equation:

Dispersion %! = CO uptake jμmol

gk

Loading l1.16µmolg

o100%

The BET surface areas of the catalysts were obtained from their nitrogen adsorption

isotherms carried out in a volumetric adsorption unit (Micromeritics, ASAP 2010). The catalysts

were degassed at 120 oC in vacuum for 6 h prior to the measurements. Adsorption at liquid

nitrogen temperature was performed using a value of 0.162 nm2 as the cross-sectional area of the

N2 molecule.

XRD measurements were made with a PANalytical X’pert Pro powder diffractometer

operated at 45 kV and 40 mA, using Cu Kα monochromatized radiation (λ = 0.154178 nm).

Crystallite size was calculated using the Scherrer equation Dc = pq@3%4 r! in which K is the shape

68

factor with a value of 0.9, λ is the X-ray wavelength, β is the line broadening at half the

maximum intensity in radians (accounting for instrumental broadening of 0.1o), and θ is the

Bragg angle.

X-ray photoelectron spectra were collected using a physical electronics PHI 5700

spectrometer with non-monochromatic Al Kα radiation (300 W, 15 kV, and 1486.6 eV) with a

multi-channel detector. Spectra of pelletized samples were recorded in the constant pass energy

mode at 29.35 eV, using a 720 µm diameter analysis area. Charge referencing was measured

against adventitious carbon (C 1s at 284.8 eV). A PHI ACCESS ESCA – V6.0 F software

package was used for acquisition and data analysis. A Shirley-type background was subtracted

from the signals. Recorded spectra were always fitted using Gaussian – Lorentzian curves in

order to determine the binding energies of different element core levels more accurately.

Reduced and spent catalysts were stored in sealed vials with an inert solvent. Prior to XPS

measurement in the analysis chamber, the solvent was evaporated off the catalyst sample in a dry

box under a N2 flow.

3.2.2. Reactivity studies

Hydrodeoxygenation of 2-methyltetrahydrofuran (2-MTHF) was carried out in a packed

bed reactor as a function of temperature at atmospheric pressure. Quantities of catalysts loaded

into the reactor were equivalent to 30 µmol of active-surface metal atoms (from CO uptake

measurements). The reactivity tests were conducted as a function of temperature at a space

velocity (GHSV) of 8000 h-1

corresponding to a contact time of 6.7 s. The space velocity and

contact time were respectively defined by the following equations:

69

GHSV = Feed volume flow rate

Volume of the catalyst bed in which feed volume flow rate is 12000 [cm

3 (NTP) h

-1]

and volume of the catalyst bed is 1.5 cm3.

Contact time = Quantity of active sites of catalyst loaded

Reactant molar flow rate

The catalysts were pretreated at the same conditions as for the chemisorption

measurements followed by cooling to 350 oC. Then, a flow of H2 [200 cm

3 (NTP) min

-1]

through a saturator containing 2-MTHF was introduced to the reactor. The saturator was kept at

0 oC and contained 95 volume % of 2-MTHF and 5% of heptane as an internal standard. Antoine

equation was used to obtain the vapour pressure of the components of the liquid mixture and

Raoult’s law was used to give the gas-phase concentration in the feed. Antoine equation is as

follows: logP4L' = A − BT+C where P

sat is the vapour pressure in mmHg, T is the bubbler

temperature (0 oC) and the coefficients A, B, and C are 7.13891, 1339.48, and 234.353

respectively for 2-MTHF and 7.04605, 1341,89, and 223.733 respectively for heptane [33].

After unit conversion, the saturated vapour pressures of 2-MTHF and heptane were 0.035 atm

and 0.015 atm respectively. Using Raoult’s law, the vapour pressure of the liquid mixture was

expressed as PULt = P<�uDvw4L' x<�uDvw + PYKt'L#K4L' xYKt'L#K where x is molar fraction of each

component in the liquid. The vapour pressure of the liquid has a value of 0.034 atm, thus, at

atmospheric pressure, the gas-phase concentration of 2-MTHF in the feed stream is 3.2 molar %.

The catalysts were stabilized for 12 h after 2-MTHF was first introduced. Temperature was

varied as follows: 350 oC � 300

oC � 250

oC � 275

oC � 325

oC � 350

oC, with the reaction

held at each temperature for 4 h. The use of a high temperature at the onset and the cycling

70

down to a low temperature and back up minimizes potential changes in the catalyst and can

indicate if deactivation has occurred. Starting at a low temperature and simply going up can

change the catalyst and give unreliable data. The catalysts showed good stability over the 48 h

of reactivity testing. The reaction products were monitored by an online gas chromatograph (SRI

8610B) equipped with an HP-1 100m x 0.25mm capillary column and a flame ionization

detector. The reactants and products were identified by comparing their retention times with

those of commercial standards and confirmed by gas chromatography – mass spectrometry (GC-

MS) (Hewlett – Packard, 5890-5927A). Results from the reactivity as a function of temperature

were used to calculate turnover frequencies.

Contact time studies for nickel phosphide and tungsten phosphide supported on silica were

conducted in a similar manner. An amount of catalyst corresponding to 5 µmol active sites was

used for each test. The experiment was carried out at a constant temperature (300 oC) while the

feed flow rate was varied so as to achieve contact times in the range of 0.3 - 7s. Contact time

studies provide information on intermediate species and reaction steps involved in the

deoxygenation of 2-methyltetrahydrofuran on the phosphides.

3.3. Results and Discussion

3.3.1. Temperature-programmed reduction

Temperature-programmed reduction (TPR) experiments were carried out to understand the

phenomena involved in the reduction process and to determine the optimum reduction conditions

for large scale catalyst preparation. Fig. 3.1 shows water (m/z = 18) and phosphine (m/z = 34)

71

evolution during the reduction of the precursors. Throughout this paper, catalysts prepared by

the phosphate method are denoted with (A) and those prepared by the phosphite method are

denoted with (I).

Fig. 3.1. Temperature-programmed reduction of transition metal phosphide catalysts

supported on silica; heating rate of 2oC/min, H2 flow rate of 100ccm/g of catalyst. a) Mass

18 signal, phosphite precursors, b) Mass 18 signal, phosphate precursors, c) Mass 34 (PH3)

signal, phosphite precursors, d) Mass 34 (PH3) signal, phosphate precursors.

The details of the synthesis of transition metal phosphides by H2-TPR depend on the metal.

In all cases it is likely that the partly or completely reduced transition metal catalyzes the

reduction of the phosphite or phosphate components to form phosphine (PH3) that itself reacts

with the reduced metal to form the transition metal phosphide [34]. The ratio of phosphorus-to-

metal (P:M) clearly affects the phosphine evolution of the catalysts reduction. Supported

72

tungsten and molybdenum phosphate precursors did not give off phosphine in the reduction

because their stoichiometric P:M ratio (1:1) dictated that all phosphorous would be consumed.

On the other hand, phosphine was released when the use of M:P ratio was 1:2. A P:M ratio equal

to 2 is required in the phosphite method in order to form Ni(HPO3H)2 and Co(HPO3H)2

precursors. The same ratio was used for iron, tungsten and molybdenum precursors for

comparative purposes. TPR profiles will be discussed over iron group phosphides (Ni2P, CoP,

and FeP) and then group 6 metal phosphides (MoP and WP).

The reduction profile for FeP/SiO2 using the phosphite precursor (Fig. 3.1a) shows a broad

signal composed of three unresolved water peaks (480, 570, 760 oC), while that using phosphate

precursor (Fig. 3.1b) also has three overlapping water peaks but at higher temperature (600, 700,

770 oC). The temperatures of the latter are higher than those reported earlier (650

oC) for the

reduction of Fe2PO4/SiO2 likely due to the use of a heating rate (2 oC min

-1) double the previous

(1 oC min

-1) [22]. In the earlier study the multiple peaks were ascribed to the presence of iron

oxide intermixed with the phosphate in the precursor, and this is likely the case here. The

formation of PH3 occurred mostly simultaneously with the reduction of the phosphite (Fig. 3.1c)

and phosphate (Fig. 3.1d), although by itself phosphate is not reducible except at high

temperature [35]. This indicates that reduced Fe is catalyzing its reduction. Overall, the

reduction of the phosphite precursor begins at a lower temperature, but finishes at approximately

the same temperature.

The reduction profile for CoP using the phosphite precursor (Fig. 3.1a) shows an initial

water peak (410 oC) followed by a slowly-developing broad feature with a high temperature

maximum (700 oC), while that using the phosphate precursor lacks the initial peak and consists

of a broad twin-peaked signal (610, 700 oC). In previous work with the phosphate precursor a

73

narrower cluster of peaks (560 oC) was observed [22]. The production of PH3 during the

reduction of the phosphite precursor (Fig. 3.1c) shows four peaks that do not correspond clearly

to stages in the overall reduction trace. In contrast, the formation of PH3 during the reduction of

the phosphate precursor (Fig. 3.1d) follows the reduction process, indicating that metal reduction

and PH3 formation occurred simultaneously. In summary for CoP the reduction of the phosphite

occurs at similar temperatures to the phosphate.

The reduction profile for Ni2P using the phosphite precursor (Fig. 3.1a) shows an initial

small water peak (420 oC) followed by a larger peak at a moderate temperature (560

oC), while

that using the phosphate precursor shows a single peak (590 oC). Results from the phosphite

method match well with previous work [23]. Previous work with the phosphate method also

showed a single reduction peak but at a lower temperature (530 oC) – an outcome of the slower

heating rate (1 oC min

-1) [17]. Studies of calcined precursors with different Ni/P ratios supported

on different types of silica clearly showed a trend that the reduction temperature of Ni increases

with increasing P content [17, 34]. This is consistent with the results obtained here since the

sample using the phosphate method has excess phosphorous (P:Ni = 2). Another study that used

in situ X-ray diffraction to follow the reduction of a phosphorous-poor Ni composition showed

that Ni was reduced at low temperatures (400 oC), and subsequent phosphidation did not occur

until moderate temperatures (550 oC) [36]. The moderate temperature is similar to the results

here. The PH3 formation during reduction of the phosphate (Fig. 3.1d) follows the overall

reduction process, with attenuation at higher temperature again due to the formation of the

phosphide. In general, Ni2P/SiO2 by the phosphite method was formed at a slightly lower

temperature than the catalyst from the phosphate method.

74

The reduction profile for MoP/SiO2 using the phosphite precursor (Fig. 3.1a) shows a small

water peak at 350 oC followed by a large peak at 560

oC. On the other hand, for the phosphate

precursor (Fig. 3.1b), the reduction of Mo occurred through an intermediate stage (375 oC)

followed by simultaneous reduction and phosphidation (520 oC), in agreement with a previous

study [18]. Previous research has established that the lower temperature water peak was due to

the reduction of MoO3 to MoO2 and that the higher temperature peak was due to the further

reduction of both molybdenum and phosphorus to form MoP [37]. For the phosphite method, the

first small PH3 peak at 450 oC is at the start of a large water peak, followed by a broad figure at

625 oC (Fig. 3.1c). The peaks do not correspond clearly to stages of MoO3 and MoO2 reduction

indicating that metal phosphidation and PH3 formation take place simultaneously. The

MoP/SiO2 phosphate precursor does not have excess phosphorus and no phosphine is detected

during the reduction, indicating that all phosphine produced was consumed for phosphidation.

Finally, the reduction profile for WP using the phosphite precursor (Fig. 3.1a) exhibits a

major water peak at 580 oC with a small shoulder at 500

oC while using the phosphate precursor

(Fig. 3.1b) the profile shows a single peak at a slightly higher temperature (600 oC). Though

there were no previous studies of the preparation of WP/SiO2 by the phosphite method, the

results for the phosphate method agree well with the previous work [20]. The TPR profile of WP

does not indicate reduction of WO3 into WO2 at similar temperatures to those found in MoP

synthesis. At higher temperatures (~ 600 oC) the tungsten precursor was reduced to W

0 which

then reacted with phosphine to form WP [20]. Phosphine (Fig. 3.1c, 1d) displays a PH3

evolution peak from the phosphite method but very little PH3 is detected from the phosphate

method. The phosphorous to metal (P:M) ratio is the key to explain this result. The WP/SiO2 (I)

precursor has a 2:1 P:M ratio and thus the excess phosphorus is released during the reduction,

75

while the WP/SiO2 (A) precursor has a 1:1 P:M ratio and there is no excess phosphorus to

release. This may also imply that all the phosphine produced was immediately consumed to

make tungsten phosphide.

In conclusion, cobalt and iron phosphides from either type of precursor were formed at

similar temperatures – CoP at 700 oC and FeP at 760

oC. The phosphite precursor of nickel

phosphide was reduced at a slightly lower temperature (560 oC) than the phosphate precursor

(590 oC). Similarly, WP (I) was formed at 580

oC and WP (A) at 600

oC. Molybdenum

phosphide did not follow the trend; MoP (I) was formed at 560 oC, whereas MoP (A) was

produced at a lower temperature (520 oC). The highest temperature at which water was detected

by the mass spectrometer plus twenty degree was chosen to be the reduction temperature in the

synthesis of larger quantities of catalysts. The materials were held at these temperatures for 2 h

to ensure complete reduction.

3.3.2. BET surface area

Table 3.2 summarizes characterization results on BET surface areas, CO chemisorption,

crystallite size, and dispersion (D) for all catalysts. The BET surface areas of the phosphides

range from 149 to 217 m2

g-1

. As reported before, supports with a low acidity such as silica or

hexagonal mesoporous silica favour the synthesis of transition metal phosphides due to the weak

interaction between support and precursor [21, 23]. It is likely that part of the pores were

partially blocked by the presence of the phosphides and the phosphorus excess on the surface,

causing a reduction in surface area. Nevertheless, all synthesized catalysts show a high surface

area, indicating a high dispersion of the active phase on the support.

76

Table 3.2. Characterization results for the supported phosphide catalysts. (I) – catalysts

made by the phosphite method, (A) – catalysts made by the phosphate method

Samples

Reduction

temperature

(oC)

BET Surface

Area (m2/g)

CO uptake

(µmol/g)

Crystallite

size (nm)

Dispersion

(%)

Ni2P (I) 580 217 137 - 12

Ni2P (A) 610 177 115 15 10

CoP (I) 720 197 138 29 12

CoP (A) 720 158 56 32 5

WP (I) 600 178 37 19 3

WP (A) 620 149 42 - 4

MoP (I) 580 220 111 5 10

MoP (A) 540 174 201 - 17

FeP (I) 780 194 102 22 9

FeP (A) 780 185 15 29 1

Pd/Al2O3 380 80 413 4 35

As expected, all catalysts from the phosphite method show higher surface areas than those

from the phosphate method. This is because the phosphite method eliminates the high-

temperature calcination step, which is used in the phosphate method. Thus, the catalysts

prepared by the phosphite method are not subjected to a “double sintering” (from calcination and

reduction) that reduces the surface area.

3.3.3. CO chemisorption

The chemisorption of CO provides an estimate of the number of active metal sites on the

surface of the catalysts. Infrared studies of CO adsorption on Ni2P [38] and MoP [39] show that

CO adsorption is mostly linear on metal sites; therefore it is reasonable to assume a

stoichiometry of one CO molecule per surface metal atom. The measured CO adsorption

capacities are presented in Table 3.2.

77

The CO chemisorption values of the different samples vary over a wide range from 15 to

200 µmol g-1

. The results of catalysts from the phosphate method are close to those reported

previously [26]. The phosphite method shows higher CO uptake than the phosphate method for

the iron group transition metal phosphides – Ni2P/SiO2, CoP/SiO2, and FeP/SiO2. The

significantly low value of FeP/SiO2 from the phosphate method is possibly due to blockage of

active metal sites by species that prevents adsorption. The phosphate precursor has phosphorus

in the high +5 oxidation state so is not easily reducible except at high temperatures [22, 24]. Iron

phosphate itself also requires high reduction temperature to convert to the phosphide [22]. It is

likely that excess phosphorus and unreduced phosphate might have remained on the surface of

the catalyst even after reduction and blocked the metal active sites. WP/SiO2 catalysts present

similar values for both synthesis methods. Meanwhile, MoP/SiO2 from the phosphate method

shows a higher CO chemisorption than the phosphite method. There is a relation between the

crystallite size (calculated from the XRD data) and CO uptake. For each pair of catalysts the

smaller crystallite size corresponds to the larger value of CO uptake, which is as expected due to

the higher dispersion of smaller catalyst particles on the support. The average CO uptake values

of the catalysts follow the order of: MoP/SiO2 > Ni2P/SiO2 > CoP/SiO2 > FeP/SiO2 > WP/SiO2.

The dispersion was calculated from CO chemisorption and the known loading of the samples,

and ranged from 1% to 17%.

3.3.4. X-ray diffraction

X-ray diffraction (XRD) data were collected to identify the crystalline phases in the

catalysts (Fig. 3.2). All catalysts show a broad diffraction line at 2θ (o) = 23-25

o, typical of

78

amorphous silica. Except for Ni2P/SiO2, catalysts made from phosphite precursors show clearer

diffraction patterns.

20 30 40 50 60 70 80

FeP/SiO2

PDF 00-039-0809

2θ (deg)

PDF 00-024-0771

MoP/SiO2

PDF 00-029-1364

WP/SiO2

CoP/SiO2

Phosphite method

Phosphate method

PDF 00-003-0953

PDF 00-029-0497

Ni2P/SiO

2

Inte

nsity / a

.u.

Fig. 3.2. X-ray diffraction patterns of supported phosphide catalysts

The XRD features for the supported iron phosphide catalysts by both methods are weak

because of fluorescence by iron; however, FeP still can be identified from the peaks at 2θ (º) =

32.7, 46.3, 47.1 and 48.3º (PDF 00-039-0809). This result is consistent with previous findings

for FeP synthesis, which in addition, showed that the synthesis requires an excess amount of

phosphorus in the precursor [22]. The XRD pattern is slightly more defined for the phosphite

79

method. A small amount of Fe2P was also detected with a feature at 2θ (º) = 40.3º (PDF 01-076-

0089).

For CoP/SiO2, though having different TPR profiles, the formation of CoP in both methods

is confirmed with clear signature peaks at 2θ (º) = 31.6, 46.2, and 48.1o (PDF 00-029-0497).

Earlier research on Co2P and CoP has established that CoP exhibits a higher activity and stability

than Co2P in the hydrotreating processes [40].

Nickel phosphide catalyst from the phosphate precursor shows the characteristic peaks of

Ni2P at 2θ (º) = 40.7, 44.5, 47.4, and 54.2o (PDF 00-003-0953). In contrast, the diffraction pattern

of nickel phosphide from the phosphite method only shows a weak, broad feature at 40.7o,

reflecting low crystallinity and small particle size probably as a consequence of the milder

preparation conditions. Using this method, small particles can be formed even with higher

loading up to (10 wt%) [24].

The MoP/SiO2 (I) shows broad peaks for MoP at 2θ (º) = 32.2 and 43.1o, whereas, the

MoP/SiO2 (A) does not present any peaks, probably because the lower reduction temperature

(540 oC) produced poorly crystallized phases. The MoP/SiO2 (I) sample and a previously

reported MoP/SiO2 (A) synthesis [18] which displayed XRD peaks used a higher temperature

(580 o

C). Other phases such as metallic Mo were not detected [37].

The XRD profile for WP/SiO2 (I) shows sharp peaks, indicating a high crystallinity WP

phase. On the other hand, though synthesized at higher reduction temperature, the WP/SiO2 (A)

only shows slight features for WP at 31.1, 43.2, and 44.6o. Apparently, it is more difficult to

form the WP phase from the phosphate precursor, which starts at a higher oxidation state.

80

The particle size of the transition metal phosphides were calculated using the Scherrer

equation for catalysts with substantive XRD patterns. CoP/SiO2 (A and I) had the largest

crystallite size followed by FeP/SiO2 (A and I), Ni2P/SiO2 (A), and MoP/SiO2 (I) (Table 3.2).

3.3.5. X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS) experiments were carried out on the most active

catalysts from each category: Ni2P/SiO2 for the iron group phosphides (Fig. 3.3) and WP/SiO2

for the group 6 phosphides (Fig. 4). Table 3 lists the binding energies for Ni 2p3/2, W 4f7/2 and P

2p3/2 core levels and the observed phosphorous to metal atomic ratios. Ni 2p and W 4f core-level

spectra were similar for both methods, although WP (I) and WP (A) were synthesized with

different molar ratios. Calculated surface atomic ratio shows similar results for Ni2P catalysts

while on WP (I) there was twice as much phosphorous as on WP (A).

Table 3.3. Spectral parameters for Ni2P/SiO2 and WP/SiO2 catalysts

Sample

Binding Energy (eV) Surface

atomic ratio Ni 2p3/2 P 2p3/2

Ni2+

Niᵟ+

/Ni0 HPO3H

- PO4

3- Ni2P P/Ni

Ni2P (I) 856.6 853.1 133.6 134.9 128.6 1.7

Ni2P (A) 856.8 853.2 - 135.0 128.6 2.1

W 4f7/2 P 2p3/2

W6+

Wᵟ+

/W0 HPO3H

- PO4

3- WP P/W

WP (I) 36.3 31.1 133.7 135.0 128.7 3.4

WP (A) 36.4 31.3 - 134.7 128.6 1.7

Ni 2p core-level spectra, for both catalysts, involve two contributions (Fig. 3.3a). The first

one is centered at 852.6 eV [24, 41, 42] and is assigned to Niδ+

in the Ni2P phase. The second

one is between 856.6 and 856.8 eV, and corresponds to Ni2+

ions interacting possibly with the

81

unreduced (HPO3H-) or PO3

-/PO4

3- anions as a consequence of a superficial passivation.

Moreover, a broad shake-up satellite signal at approximately 6.0 eV above the Ni2+

species is

observed due to the presence of divalent species [24, 40, 41]. The reduced Ni species from the

Ni2P phase has a binding energy that is very close to that of Ni metal (852.5-852.9 eV) [43], so

its presence cannot be ruled out in both catalysts. The similarity of the XPS features for

Ni2P/SiO2 via both methods and the confirmation by XRD of the presence of Ni2P in the catalyst

made by the phosphate method, and the catalytic activity to be presented later provide reasonable

support for the presence of Ni2P on the catalyst made by phosphite method even though no peaks

were observed in the XRD profile of Ni2P/SiO2 (I).

885 880 875 870 865 860 855 850 138 136 134 132 130 128

Ni2P (I)

Ni 2p core level spectra

Ni2P (A)

Inte

nsity /

a.u

.

Binding Energy / eV

P 2p core level spectra

Binding Energy / eV

Ni2P (I)

Ni2P (A)

Fig. 3.3. Ni 2p and P 2p core-level spectra for silica supported nickel phosphide catalysts

82

The P 2p core-level spectrum for Ni2P (A), phosphate method, shows two contributions

(Fig. 3.3). The band at low binding energy values, 129.3 eV, is assigned to phosphorus in the

form of metal phosphide, Ni2P [24], while the band located at 135.0 eV is due to PO43-

species

which appears as a consequence of the superficial passivation of phosphide species. The P 2p

core-level spectrum for Ni2P (I), phosphite method, shows an additional band assigned to

unreduced phosphite species at 133.4 eV.

The W 4f photoemission spectra of the tungsten phosphides prepared by both methods are

shown in Fig. 3.4. The decomposition of the spectra reveals several overlapping bands (Fig.

3.4). The first contribution is located at 31.2 eV and is ascribed to Wᵟ+

or W0 species as

previously reported [44] and is akin to that observed in tungsten phosphide [20]. The second

contribution located at about 36.3 eV is assigned to W6+

species [45].

42 40 38 36 34 32 30 28 138 136 134 132 130 128 126

WP(I)

WP(A)

Inte

nsity /

a.u

.

Binding Energy / eV

W 4f core level spectra P 2p core level spectra

Binding Energy / eV

WP(A)

WP(I)

Fig. 3.4. W 4f and P 2p core-level spectra for silica supported tungsten phosphide catalysts

83

The P 2p spectra (Fig. 3.4) for tungsten phosphide catalysts are similar to those obtained

for Ni2P catalysts. The contributions are due to the phosphide phase at 128.7 eV and phosphate

groups at 135.0 eV. The WP (I) catalyst also reveals the presence of phosphite (133.6 eV) due to

the phosphorous excess present on this sample. The matching XPS figures of catalysts from both

methods indicate that there is a WP phase in the bulk of the WP (A) particles although the XRD

pattern did not show strong peaks.

3.3.6. Reactivity in hydrodeoxygenation (HDO) of 2-methyltetrahydrofuran (2-MTHF)

3.3.6.a)Reactivity as a function of temperature

The catalytic activity of these phosphides were evaluated for the HDO of 2-MTHF on a

basis of 30 µmol active sites as measured by CO chemisorption. The gas phase feed contained

3.2 % of 2-MTHF in a H2 stream with a H2 to feed ratio of 30:1. The catalytic activities of all

phosphides at different temperatures are presented as turnover frequencies (TOF) in Fig. 3.5. All

catalysts have higher conversion at higher temperature as expected. Values of TOF together with

the corresponding conversions at 275 oC and 350

oC are reported in Table 3.4.

84

Table 3.4. Conversions and apparent activation energies of 2-MTHF HDO on transition

metal phosphides

Sample Total Conversion % at

Temperature oC

Apparent

activation

energy Ea

(kJ/mol)

TOF (s-1

)

at 275oC

TOF (s-1

)

at 350oC

250 275 300 325 350

Ni2P (I) 4 12 43 63 76 114 0.020 0.12

Ni2P (A) 6 20 63 91 97 119 0.032 0.15

CoP (I) 1 4 10 35 66 108 0.0058 0.11

CoP (A) 0.5 2 8 17 40 137 0.0039 0.06

WP (I) 6 16 41 56 65 88 0.025 0.096

WP (A) 4 11 29 48 64 101 0.016 0.1

MoP (I) 2 5 21 32 63 113 0.0084 0.1

MoP (A) 2 6 17 30 31 109 0.0091 0.05

FeP (I) - - 1 3 2 116 - 0.0035

FeP (A) 0.5 1.4 5 14 24 117 0.0023 0.038

Pd/Al2O3 0.4 1 2 4 5 77 0.0016 0.0074

At high temperature (350 oC) the phosphide catalysts prepared by the phosphite method

perform with little difference at TOF ~ 0.1 s-1

with an order of Ni2P/SiO2 > CoP/SiO2 =

MoP/SiO2 ~ WP/SiO2. The only exception is iron phosphide (I) which exhibits an extremely low

activity - only 12 molecules of substrate got converted per active site per hour. For the

phosphate method, a more distinct order was observed: Ni2P/SiO2 > WP/SiO2 > CoP/SiO2 ~

MoP/SiO2 > FeP/SiO2. In both methods, Ni2P is the most effective catalyst, as observed in many

HDS, HDN, and HDO studies [15, 22, 26]. A comparison within each pair of catalysts shows

that the TOF values of CoP and MoP using the phosphite method are higher than those of the

phosphate method. Meanwhile, for Ni2P, WP, and FeP, catalysts prepared by the phosphate

method are better than those made by the phosphite method, though the extents of differences

vary. The TOF of Ni2P (A) is about 1.4 times that of Ni2P (I), WP (A) is practically the same as

WP (I), while FeP (A) is ten-fold more active than FeP (I).

85

250 275 300 325 350

0.00

0.05

0.10

0.15(a) Phosphite Method

Ni2P

CoP

WP

MoP

FeP

Pd/Al2O

3

Temperature oC

1.6x10-3

1.7x10-3

1.8x10-3

1.9x10-3

-8

-6

-4

-2

Turn

ove

r F

req

uency s

-1Ln (T

urn

ove

r Fre

quency s

-1)

250 275 300 325 350

0.00

0.05

0.10

0.15 (b) Phosphate Method

Ni2P

CoP

WP

MoP

FeP

Pd/Al2O

3

1.6x10-3

1.7x10-3

1.8x10-3

1.9x10-3

-8

-6

-4

-2

Y A

xis T

itle

Temperature-1 K

-1

Fig. 3.5. Turnover frequency of 2-MTHF on transition metal phosphides

At lower temperature (275 oC), in both methods, supported FeP still shows the lowest

activity as at higher temperature. Iron phosphides have been shown to be prone to coking and

deactivation [22, 26]. Though most of previous studies focused on metal rich Fe2P/SiO2, similar

weak activity and easy deactivation were also observed on FeP/SiO2 in this HDO study of 2-

MTHF. The order of activity of the catalysts prepared by the phosphite method is: WP > Ni2P >

MoP > CoP > FeP and that by the phosphate method is: Ni2P > WP > MoP > CoP > FeP.

Turnover frequencies within each rank are in the same order; for instance, the average TOF

values are the following: 1st (WP (I) and Ni2P (A)): 0.03 s

-1, 2

nd (Ni2P (I) and WP (A)): 0.02 s

-1,

3rd

(MoP): 0.009 s-1

, and 4th

(CoP): 0.005 s-1

. Overall, the differences in the TOF obtained from

86

using the phosphite or phosphate preparation methods are not extraordinarily large, indicating

that the final materials are similar. However, as will be seen, the phosphite method can result in

the deposition of excess phosphorus on the surface, which can affect the HDO reaction pathway.

A comparison of conversions for the catalysts prepared by the two preparation methods is

given in Fig. 3.6. The graph allows visualization of the activity order among the catalysts. In

general, phosphides from the phosphite method have higher conversions than those from the

phosphate method, except for the supported Ni2P and FeP. The catalysts’ average total

conversions have the following order: Ni2P/SiO2 (16%) > WP/SiO2 (13.5%) > MoP/SiO2 (5.5%)

> CoP/SiO2 (3%) > FeP/SiO2 (0.7%). Supported Ni2P, WP, and MoP catalysts show a steady

trend of high HDO conversion (90-100%) in the temperature range 250 – 350 oC. Regardless of

preparation method, Ni2P/SiO2 and WP/SiO2, representatives of iron group and group 6

phosphides, stand out in their performance. For this reason the samples were chosen for further

study.

250 275 300 325 3500

25

50

75

100

Se

lectiv

ity to

wa

rds

HD

O p

rod

ucts

/ %

Se

lectivity t

ow

ard

s

HD

O p

rod

ucts

/ %

Con

ve

rsio

n / %

Ni2P (I)

WP (I)

MoP (I)

CoP (I)

FeP (I)

Pd/Al2O

3

Co

nve

rsio

n /

%

250 275 300 325 3500

25

50

75

100 Ni

2P (A)

WP (A)

MoP (A)

CoP (A)

FeP (A)

Pd/Al2O

3

250 275 300 325 350

60

80

100

Temperature / oC

Temperature / oC

250 275 300 325 350

60

80

100

Fig. 3.6. Total conversion and selectivity towards HDO products of 2-MTHF reaction on

transition metal phosphides

87

Table 3.4 shows the total conversions at various temperatures together with the apparent

activation energy Ea and Fig. 3.5 exhibits the TOF results in Arrhenius plots. The apparent

activation energy was calculated from the TOF values at lower temperatures (250-300 oC), using

the Arrhenius equation. Iron phosphide from the phosphite precursor did not have any activity at

275 oC and showed deactivation at 350

oC, thus, its apparent activation energy was calculated in

the range of 300 - 325 oC. Activation energies of the supported catalysts span the range of 88

kJ/mol - 137 kJ/mol. These values are lower than the C-O bond dissociation energy (360 kJ/mol),

suggesting that direct CO bond scission did not occur. Oxygen removal may have occurred by a

protonation of the oxygen first, followed by a hydrolysis.

Ni2P catalysts from both methods show similar activation energy. The same was found for

FeP. For CoP and WP, catalysts from the phosphite method have lower activation energy than

those from the phosphite method, while the opposite was found for MoP/SiO2 (Table 3.4). The

differences in activation energy are not great, except possibly for the WP samples. The apparent

activation energy calculated by the Arrhenius equation only provides an indication of the barrier

for the overall chemical reaction, which might contain positive contributions from activated steps

but negative contributions from equilibrated steps. Thus, when a multiple step sequence is

involved, the order of activation energies is not directly related to a key reaction step unless the

same mechanism is involved with all catalysts. As will be seen, this is not the case here.

The reactivity data for the commercial 5 wt% Pd/Al2O3 catalyst are reported in Fig. 3.5,

Fig. 3.6, and Table 3.4. In general, the palladium catalyst was substantially out-performed by

the phosphide catalysts with the only exception of FeP (I). As temperature increased from 250 to

350 oC, the total conversion of 2-MTHF on the Pd catalyst increased from 0.4 to 5%. These low

88

conversions may be due to coking and/or poisoning. A previous study [26] also reported a low

performance of Pd/Al2O3 catalyst in the HDO of guaiacol. In the study the palladium catalyst

showed a strong preference towards cleavage of the methyl group from guaiacol, producing

catechol – a coking-inducing product. Product analysis for HDO of 2-MTHF (Fig. 3.9) consisted

of a large proportion of a C4 mixture (30 - 45%), indicating a similar removal of a methyl group

from 2-MTHF. Carbon residues might have formed a coke layer on the palladium catalyst,

blocked the active sites, and thus resulted in the low conversions.

Product selectivity was calculated by the following equation:

Selectivity =molofproduct

molof2-MTHFreacted100%

Table 3.5 reports total conversions, HDO conversions, and HDO product selectivity of all

the catalysts in the region of moderate temperature and conversion. Special exception was made

for FeP/SiO2 because no product was detected for FeP/SiO2 (I) at 275 oC, so values of HDO

product selectivity for the iron phosphides at 300 oC are listed. Fig. 3.7-3.9 provide a visual

presentation of specific products of the reaction on the phosphides. Based on the HDO products,

the catalysts can be categorized in two groups. The first group consists of supported Ni2P and

CoP with C4 and C5 n-alkanes (such as butane and pentane) and 2-pentanone as predominant

products (Fig. 3.7). The second group comprises supported group 6 metal phosphides (WP and

MoP) and FeP with the main products being unsaturated C5 hydrocarbons (pentenes and

pentadienes), pentane, and C4 hydrocarbons (Fig. 3.8, Fig. 3.9). Though total conversions of 2-

MTHF vary over a wide range (1%-16%), all phosphides demonstrated a strong preference to

HDO products (>74%). The order of average HDO conversions is as follows: MoP/SiO2 >

FeP/SiO2 > WP/SiO2 > Ni2P/SiO2 > CoP/SiO2.

89

Table 3.5. Product selectivity for HDO of 2-MTHF on transition metal phosphides at 275 oC

*Conversions of FeP/SiO2 were at 300oC

Sample

Total

conversion %

(at 275oC)

Conversion

to HDO

products %

HDO products selectivity %

Pentane Pentenes Pentadienes C4

Mixture

Ni2P (I) 12 85 67 2 1 15

Ni2P (A) 20 81 61 0 0 20

CoP (I) 4 74 43 2 0 29

CoP (A) 2 80 40 15 0 25

WP (I) 16 78 0 11 65 2

WP (A) 11 95 3 68 23 1

MoP (I) 5 95 14 70 7 4

MoP (A) 6 90 2 44 39 5

FeP (I)* 1 91 5 64 5 15

FeP (A)* 5 86 15 55 0 16

Pd/Al2O3 1 72 17 12 0 43

Butane

1-pentene

pentane

t-2-pentene

c-2-pentene

methylfuran

2-pentanone

0

20

40

60

80

Pro

du

ct S

ele

ctivity / %

CoP/SiO2 Total Conversion 3%

Phosphite Method

Phosphate Method

0

20

40

60

80

Ni2P/SiO

2 Total Conversion 5%

Phosphite Method

Phosphate Method

Fig. 3.7. Product selectivity of 2-MTHF reaction on Ni2P/SiO2 and CoP/SiO2 at a total

conversion of around 5%

90

The iron group metal (Ni, Co, and Fe) phosphides were expected to give similar HDO

product distributions, yet only Ni2P and CoP followed this trend (Fig. 3.7). The products of 2-

MTHF HDO on Ni2P/SiO2 were mainly pentane (61-67%), butane (15-20%), 2-pentanone (5-

8%), and very little alkenes (0-3%). CoP/SiO2 produced less pentane (40-43%) and more 2-

pentanone (10-16%) than Ni2P/SiO2. These products suggest that besides oxygen removal there

were also dealkylation and dehydrogenation occurring and that 2-pentanol may have been an

intermediate in the HDO mechanism, though not observed in the products stream. Ni2P/SiO2

performed better than CoP/SiO2 with higher total conversions (>12%), higher HDO conversions

(>81%), and higher selectivity to pentane (>61%). The Ni2P/SiO2 prepared by the phosphite

method was better than the phosphate method in terms of higher HDO conversions and better

yield toward pentane. In comparison, CoP/SiO2 (A) was better than CoP/SiO2 (I) because of the

higher total HDO product yields.

91

Butane

1,4-pentadiene

1-pentene

pentane

t-2-pentene

c-2-pentene

t-1,3-pentadiene

furan

c-1,3-pentadiene

methylfuran

2-pentanone

n-pentanal

0

20

40

Pro

du

ct S

ele

ctivity / %

MoP/SiO2 Total Conversion 5%

Phosphite Method

Phosphate Method

0

20

40

WP/SiO2 Total Conversion 5%

Phosphite Method Phosphate Method

Fig. 3.8. Product selectivity of 2-MTHF reaction on WP/SiO2 and MoP/SiO2 at total

conversion of 5%

Group 6 metal phosphides (WP/SiO2 and MoP/ SiO2) showed a higher HDO conversion

(78% to 95%) compared to iron group metal phosphides (Table 3.5). A clear preference towards

unsaturated hydrocarbons was observed (Fig. 3.8) with very little 2-pentanone and C4-mixtures

detected. This suggests that a different ring opening mechanism took place. The HDO of 2-

MTHF with tungsten phosphides (I) and (A) shows different product distributions. For WP/SiO2

(I) the formation of pentadienes was favoured, whereas for WP/SiO2 (A) pentenes were preferred

(Fig. 3.8). In general, WP/SiO2 (A) gave a notably higher selectivity towards HDO (95% vs.

78%, Table 3.5). The surface metal-to-phosphorus ratio may have been responsible for this

92

sensitivity (Table 3.3). The P/W ratio was 3.4 for the phosphite method and 1.7 for the phosphate

method. This will be discussed shortly.

Both MoP/SiO2 (I) and (A) showed the highest HDO activity (although not total

conversions) among all the phosphides examined here with MoP (I) better than MoP (A) (Table

3.5). The synthesis method also seems to affect the reaction pathways of the catalysts. MoP/SiO2

(I) favoured production of pentenes and pentane while MoP/SiO2 (A) gave additional large

portions of pentadienes (Fig. 3.8). The phosphorus-to-metal ratio may again have been

responsible for this sensitivity. The P:M ratio for preparation was 2:1 for the phosphite method

and 1:1 for the phosphate method (Table 3.1).

The HDO of 2-MTHF on iron phosphide generated 5-15% of pentane, 15% of

dealkylation products, 4-6% of 2-pentanone, and 55-64% of pentenes (Fig. 3.9). The large

proportion of dealkylation products as well as the noticeable amount of 2-pentanone product

resembles the results of other iron group metal phosphides (Ni2P and CoP). Yet, more than half

of the products were alkenes, indicating the occurrence of dehydration. FeP/SiO2 (I) showed a

higher selectivity towards HDO products; however, it had a much lower overall conversion than

FeP/ SiO2 (A).

93

Butane

1,4-pentadiene

1-pentene

pentane

t-2-pentene

c-2-pentene

t-1,3-pentadiene

furan

c-1,3-pentadiene

methylfuran

2-pentanone

n-pentanal

0

10

20

30

Pro

du

ct S

ele

ctivity / %

Pd/Al2O

3 Total Conversion 5%

0

10

20

30

FeP/SiO2 Total Conversion 4%

Phosphite Method

Phosphate Method

Fig. 3.9. Product selectivity of 2-MTHF reaction on FeP/SiO2 and Pd/Al2O3 at total

conversion of 5%

Pd/Al2O3 exhibited a high propensity for the removal of the methyl group from the 2-

methyltetrahydrofuran reactant. The main products included an average of 12% pentane, 25% of

pentenes, 38% of a C4 mixture, and 16% of 2-pentanone (Fig. 3.9). Similar results have been

reported earlier on Pd/SiO2 [46]. The large amount of C4-hydrocarbons and the absence of

methane production coupled with a poor reactivity imply that coke deposition and/or self-poison

by CO by-product occurred. The earlier study also showed substantial CO formation on

supported metals of group 10 (Ni and Pt) – the same group as Pd – during HDO of 2-MTHF [47,

48]. In other studies of 2-MTHF hydrogenolysis on Pt/SiO2 it was suggested that the produced

94

CO adsorbs strongly on the Pt surface and causes self-poisoning of the catalyst [47, 48]. This

may also be the case here with Pd/Al2O3.

3.3.6.b)Reactivity as a function of contact time study

Contact time studies provide trends of product selectivity that help to identify reaction

intermediate products and determine reaction pathways. Nickel and tungsten phosphides

supported on silica have the highest HDO performance from the iron group and group 6

transition metals and therefore were chosen for study. At the moderate temperature of 300 oC

and low contact time ranging from 0.3 s to 7 s, the catalysts exhibited a low catalytic activity

(15% total conversion or less) in the kinetic-regime. Absence of internal diffusion limitations

was checked using the Weisz-Prater criterion (CWP) (see Supplementary Information) [49]. The

highest value of CWP was 0.3, satisfying the criterion (CWP < 1), suggesting no internal diffusion

limitations or concentration gradient within the catalyst. Fig. 3.10 shows selectivity of products

in the HDO of 2-MTHF on Ni2P/SiO2. The contact time study on Ni2P/SiO2 (A) was chosen as

representative in the iron group since catalysts from both synthesis methods demonstrated similar

product selectivities during the reactivity experiments as a function of temperature (Fig. 3.7,

Table 3.5). The product selectivities are presented in three groups: alkenes (1-pentene and 2-

pentenes), oxygen containing compounds, and pentane. Selectivity for alkene products (trans,

cis 2-pentene, and 1-pentene respectively) decreases with increasing contact time from 17%, 8%,

and 6% to 5%, 2.5%, and 1.5% (Fig. 3.10), indicating that these are intermediate product species.

Dehydrogenated products such as methylfuran also decrease with increasing contact time. On

the other hand, the oxygenated products, 2-pentanone and n-pentanal, show the behavior

characteristic of secondary products as they grow to a maximum (14% and 3.2% respectively)

95

and then decline. Pentane increases steadily from 28.5% to 49.5% as contact time progresses,

demonstrating the characteristic of a final product (Fig. 3.10).

0 5 10 15 20 25 30

0

25

50

75

100

0

25

50

75

100

Conversion

Pentane

Butane

0

5

10

15

0

5

10

15

Se

lectivity /

%

Contact time /s

Ni2P/SiO

2 (Phosphate Method)

2-pentanone

2-methylfuran

n-pentanal

0

5

10

15

0

5

10

15

t-2-pentene

c-2-pentene

1-pentene

Fig. 3.10. Contact time study on Ni2P/SiO2

The results suggest that more than one reaction pathways occurs on the surface of

Ni2P/SiO2. The less dominant reaction route involves dehydrogenation of 2-MTHF into a very

small amount of methyldihydrofuran as primary product and methylfuran as secondary product.

The main reaction route involves the production of alkenes, 2-pentanone and pentane. This route

consists of a ring opening process and a C-O cleavage to give alkenes as primary product, 2-

pentanone as secondary product, and pentane as final product. A consecutive reaction pathway

such as 2-MTHF � alkenes � 2-pentanone � pentane is unlikely because of the need to

oxidize alkenes to 2-pentanone in a highly reductive experimental environment. A more likely

96

possibility is that there are common intermediate species at the surface that react to form alkenes,

2-pentanone, and pentane in a sequential manner. This observation is similar to a previous study

on the conversion of ethanol on Ni2P/SiO2 where ethanol was converted to acetaldehyde as

primary product and ethylene as secondary product [50]. The contact time study results lead to

the following “rake mechanism” scheme (Scheme 3.1). A rake mechanism is one which involves

sequential steps of adsorbed species on a surface with each adsorbed species capable of

desorbing. Thus, a diagram of the scheme shows prongs like in a rake for sweeping leaves.

The results can be rationalized in the following manner. The reactant 2-MTHF first

adsorbs onto a vacant site on the catalyst surface. Nucleophilic attack by a surface hydride in

step � opens the ring to form a 1o or 2

o alkoxide intermediate. Attack by a nearby electron-rich

site on a hydrogen beta to the oxygen atom gives rise to an E2 elimination in step � to produce

the primary alkene products. The alkoxide intermediates may also go through an alpha hydride

elimination in step � to form adsorbed carbonyl species with hydrogen transfer to vacant active

sites. The carbonyl species can desorb as 2-pentanone or n-pentanal or can react further through

a hydrogen transfer in step � to form adsorbed alcohol species. These can then undergo another

hydrogen transfer in step � and release pentane.

97

O

+*

O

*

+

O

*

O

*

+ OH

*

O

*

O

O

Alkenes 2-pentanone Pentane2-methyltetrahydrofuran

n-pentanal

2-methylfuran

1O

*

+

H

*

-

2O

+

*

H

H

H

*

O

*

H

*

O

*H

**

OH

**

H

*

3

4 5

OH

**

-

or

*

O

or

O

O

*

+H

*

-

or or

OH

*

or

O

* H

**or

OH

**

H

*

-

or

1

2

3 4

5

+ +

+

Scheme 3.1. HDO reaction network of 2-MTHF on Ni2P/SiO2

In contrast to nickel phosphide, the HDO of 2-MTHF on tungsten phosphides supported on

silica produced mostly unsaturated products. Interestingly, the selectivities followed different

trends between the two catalyst preparation methods. Fig. 3.11 shows the selectivity to the three

main groups of products: pentadienes, pentenes and pentane, and oxygenated compounds.

98

0 1 2 3 4 5 6 7

10

20

30

40

Se

lectiv

ity /%

WP/SiO2 (Phosphite Method)

Se

lectivity /

%

t-1,3-pentadiene

1,4-pentadiene

c-1,3-pentadiene

0 1 2 3 4 5 6 7

10

20

30

40 t-1,3-pentadiene

1,4-pentadiene

c-1,3-pentadiene

WP/SiO2 (Phosphate Method)

0 1 2 3 4 5 6 70

10

20

t-2-pentene

1-pentene

c-2-pentene

pentane

0 1 2 3 4 5 6 70

10

20 t-2-penteme

1-pentene

c-2-pentene

pentane

0 1 2 3 4 5 6 70

10

20 Furan

Conversion

Contact time /s

0 1 2 3 4 5 6 70

10

20 Furan

Conversion

1-pentanal

Contact time /s

Fig. 3.11. Contact time study on WP/SiO2 prepared by the phosphite and phosphate

methods

Clear trends were observed on WP prepared by the phosphate method. The main products

were pentadienes and pentenes. Selectivity to pentadienes decreased with increasing contact

time, from 26%, 28%, and 20% to 19%. 19%, and 11% (1,4-pentadiene, t-1,3-pentadiene, and c-

1,3-pentadiene respectively). In contrast, selectivity to pentenes increased with increasing

contact time, suggesting that a hydrogenation process follows the production of pentadienes.

Though pentane shares the same trend as with pentenes, its selectivity is extremely low

(maximum 1.6%). Furan selectivity starts at 14% then peaks at 15% before quickly going down

to 3.4%. Some pentanal is also detected though at low selectivity (< 2%) starting at 1.1% to

reach 1.6% at the end.

99

The results demonstrate that pentadienes were formed first then pentenes and pentane, the

latter produced by hydrogenation. This leads to a dominance of pentenes as final products at

high conversion at steady-state. This interpretation agrees well with the steady-state results on

supported tungsten phosphide (phosphate method) where at a 29% total conversion, HDO

products were mainly pentenes (74%) rather than pentadienes (21%) (Fig. 3.8). These

observations can be rationalized by another reaction scheme (Scheme 3.2). In order to produce

pentadienes as primary products, it is likely that 2-MTHF was bound to two active sites. The 2-

MTHF is first absorbed on a single site and then in step � a surface nucleophilic species attacks

a carbon alpha to the oxygen to form a doubly bound intermediate species. Next, nearby vacant

sites induce beta hydride elimination in step � and release pentadienes. The elimination follows

E2 mechanism and the resulting pentadienes desorb immediately, having little chance to

isomerize. Furthermore, at low residence time, the surface reaction follows kinetic regime which

allows the production of the kinetic product – 1,4-pentadiene. The following step involves

hydrogenation to form alkenes. The small amount of n-pentanal observed may have occurred by

an alpha hydride abstraction from primary alkoxide species in step �, followed by a hydrogen

transfer to detach n-pentanal from the catalyst surface.

100

Scheme 3.2. HDO reaction network of 2-MTHF on WP/SiO2 (phosphate method)

Interestingly, the contact time results from the HDO reaction on WP by the phosphite

method show almost the opposite trend from that obtained by the phosphate method (Fig. 3.11).

Pentadienes are the final products, except for c-1,3-pentadiene. The selectivity to 1,4-pentadiene

and t-1,3-pentadiene respectively increases from 9.5% and 33.5% to 27% and 40.5%, whereas c-

1,3-pentadiene selectivity starts at 21% to decrease only slightly to 19%. Meanwhile, 1-pentene

and c-2-pentene have an initial selectivity of 5% and 5.2% that later decline to 2% and 1.3%

respectively, indicating that they are primary products. The selectivity to t-2-pentene decreases

101

slightly from 4% to 2.6% during the initial period of contact time (0.4 s to 1 s) prior to a rise to

3.8%. Pentane was only produced at 0.7 s (0.75%) and 1 s (0.43%). Furan selectivity decreases

steadily from 21% to 2.5% over the contact time range of 0.4 s to 7 s. These results are in good

agreement with the steady-state findings at moderately high conversion (36%) in the reactivity

test where the products are comprised of 81% pentadienes and 15% of pentenes (Fig. 3.8).

The results on tungsten phosphide made by the phosphite method can be rationalized by a

scheme (Scheme 3.3) which has a similar starting point with the one for the phosphate method

where an intermediate species formed by step � is bound to two surface sites. However, a

difference is that hydrogen removal occurs stepwise. A single beta-hydrogen abstraction from

the intermediate coupled with a hydride transfer to release alkene products occurs as in step �.

Without a second hydride transfer to cause desorption from the catalyst surface, the doubly

bound intermediate species transforms into an alkene intermediate species in step � which can

react further via another beta hydrogen elimination to form pentadienes in step �. The

difference between the two pathways can be accounted for by the lack of readily available vacant

sites on the surface of WP/SiO2 by the phosphite method. In the phosphite preparation, the

phosphorus-to-metal ratio was 2:1, twice as much phosphorous compared to the ratio of 1:1 used

in the phosphate method. Furthermore, the phosphite method did not undergo calcination to

remove the excess surface phosphorus. XPS data showed the surface P/M ratio was 3.4/1 for the

phosphite method compared to 1.7/1 for the phosphate method. As a result, the phosphorous

residues produce a catalyst surface with less vacant sites and this leads to a step by step

consecutive dehydrogenation on WP made by the phosphite method. The findings demonstrate

that surface composition in the form of phosphorus to metal ratio can play a crucial role in the

reaction pathway adopted by the reactant.

102

Scheme 3.3. HDO reaction network of 2-MTHF on WP/SiO2 (phosphite method)

3.4. Conclusions

Nickel, cobalt, tungsten, molybdenum, and iron phosphides were synthesized by reducing

phosphite precursors and phosphate precursors. At 300 oC and 1 atm total conversions of 2-

103

methyltetrahydrofuran (2-MTHF) on supported phosphides followed the order of Ni2P > WP >

MoP > CoP > FeP > Pd/Al2O3. At 5% total conversion, the selectivity towards HDO products

had the following trends: MoP > WP > Ni2P > FeP > CoP (phosphite method) and MoP ~ WP >

FeP > Ni2P > CoP (phosphate method). In general, supported catalysts from the phosphite

method had higher surface area and comparable selectivity towards HDO products at 5% total

conversion. The main HDO products of supported iron metal phosphides group (Ni2P and CoP)

were pentane and butane, whereas those of group 6 metal phosphides and iron phosphide were

mostly pentenes and pentadienes. The commercial catalyst Pd/Al2O3 and the weaker catalyst of

the iron group – FeP/SiO2 produced mostly pentenes and C4 mixtures. The results from contact

time studies were used in the development of reaction networks for Ni2P/SiO2 and WP/SiO2 for

both preparation methods. Ni2P/SiO2 by both methods had similar selectivity profiles which

could be explained by a rake mechanism that produced pentenes as primary products, 2-

pentanone as a secondary product and pentane as a final product. On the other hand, selectivity

profiles for WP/SiO2 depended greatly on the preparation method. WP/SiO2 by the phosphate

method produced pentadienes prior to pentenes; therefore, the surface intermediate is suggested

to have been bound to two active sites to allow simultaneous dehydrogenation for the initial

pentadienes formation. WP/SiO2 by the phosphite method produced pentenes prior to

pentadienes indicating that sequential hydrogen removal took place on single sites. X-ray

photoelectron spectroscopy of the WP from the phosphite method indicated more phosphorus on

the surface than WP from the phosphate method, consistent with more single metallic sites

present in the latter.

104

3.5. Supplementary Information

Weisz – Prater criterion for internal diffusion

The Weisz – Prater criterion CWP uses measured values of the rate of reaction to determine

if internal diffusion is limiting the reaction as follows:

CWP = Actual reaction rate

A diffusion rate=−|} ~��!� ���<

.�g}�

Internal mass transfer effects can be neglected when the value of CWP is less than 1. The

parameters used for the criterion are listed in Table 3.6. The observed rates −|} ~��!� were

calculated at each contact time and listed along with the corresponding CWP values in Table 3.7-

3.9. Most values of CWP are on the order of 10-2

and the highest value is around 0.3 < 1,

indicating that internal mass transfer effects can be neglected at the conditions employed for

reactivity study in this paper.

Table 3.6. Parameters in the Weisz-Prater criterion

R Catalyst particle radius / cm 0.08

ρc (Ni2P) Solid catalyst density for Ni2P / g·cm-3

0.3

ρc (WP) Solid catalyst density for WP / g·cm-3

0.4

CAs (300oC) Gas concentration at the catalyst surface / mol·cm

-3 6.8 x10

-7

De Effective diffusivity/ cm2·s

-1 0.1

Table 3.7. The observed reaction rates and corresponding Weisz-Prater criterion for

Ni2P/SiO2

Contact time / s −|} ~��!� x 106

Observed reaction rate / mol g-1

s-1

CWP

Weisz-Prater criterion

0.39 2.09 0.059

0.56 2.28 0.064

0.75 1.66 0.047

1.13 1.84 0.052

105

2.26 1.45 0.041

6.38 7.98 0.225

9.18 9.76 0.276

10.75 8.27 0.233

30.40 3.20 0.090

Table 3.8. The observed reaction rates and corresponding Weisz-Prater criterion for

WP/SiO2 (I)

Contact time

/ s −|} ~��!� x 10

6

Observed reaction rate / mol g-1

s-1

CWP

Weisz-Prater criterion

0.34 0.35 0.013

0.39 0.42 0.016

0.47 0.46 0.017

0.73 0.47 0.018

1.10 0.48 0.018

2.03 0.46 0.017

6.84 0.64 0.024

Table 3.9. The observed reaction rates and corresponding Weisz-Prater criterion for

WP/SiO2 (A)

Contact time

/ s −|} ~��!� x 10

6

Observed reaction rate / mol g-1

s-1

CWP

Weisz-Prater criterion

0.40 0.89 0.033

0.56 0.46 0.017

0.75 0.35 0.013

1.13 0.36 0.014

2.26 0.27 0.010

3.46 0.21 0.008

7.51 0.24 0.009

106

References

1. A. A. Lappas,S. Bezergianni,I. A. Vasalos, Catal. Today, 145 (2009) 55.

2. G. W. Huber,S. Iborra,A. Corma, Chem. Rev., 106 (2006) 4044.

3. J. J. Bozell ,G. R. Petersen, Green Chem. , 12 (2010) 539.

4. J. Zakzeski,P.C.A. Bruijnincx,A.L. Jongerius,B.M. Weckhuysen, Chem. Rev., 110 (2010)

3552.

5. D. Mohan,C. U. Pittman Jr.,P. H. Steele, Energ. Fuel, 20 (2006) 848.

6. L. Zhang,C. Xu,P. Champagne, Energ. Convers. Manage., 51 (2010) 969.

7. A. Oasmaa ,S. Czernik, Appl. Catal. A: Gen., 13 (1999) 914.

8. Bridgwater, A. V., Appl. Catal. A: Gen., 116 (1994) 5.

9. Demirbas, M. F., Appl. Energ., 86 (2009) S151.

10. Furimsky, E., Appl. Catal. A: Gen., 199 (2000) 147.

11. P. M. Mortensen,J. D. Grunwaldt,P. A. Jensen,K. G. Knudsen,A. D. Jensen, Appl. Catal.

A: Gen., 407 (2011) 1.

12. A. Gutierrez,R. K. Kaila,M. L. Honkela,R. Slioor,A. O. I. Krause, Catal. Today, 147

(2009) 239.

13. A. Y. Bunch ,U. S. Ozkan, J. Catal. , 206 (2002) 177.

14. E. M. Ryymin,M. L. Honkela,R. R. Viljava,A. O. I. Krause, Catal. Today, 147 (2009)

239.

15. Oyama, S. T., J. Catal., 216 (2003) 343.

16. S. T. Oyama, T. Gott, H. Zhao, Y. -K. Lee, Catal. Today, 143 (2009) 94.

17. S. T. Oyama, X. Wang, Y. -K. Lee, K. Bando, F. G. Requejo, J. Catal., 210 (2002) 207.

18. P. Clark,X. Wang,S. T. Oyama, J. Catal., 207 (2002) 256.

19. W. Li,B. Dhandapani,S. T. Oyama, Chem. Lett. , (1998) 207.

20. P. Clark,W. Li,S. T. Oyama, J. Catal., 200 (2001) 140.

21. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, Appl. Catal.

B: Environ., 92 (2009) 100.

22. X. Wang,P. Clark,S. T. Oyama, J. Catal., 208 (2002) 321.

23. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, J. Phys.

Chem. C, 113 (2009) 17032.

24. J. A. Cecilia,A. Infantes-Molina,E. Rodríguez-Castellón,A. Jiménez-López, J. Catal., 263

(2009) 4.

25. S. T. Oyama, X. Wang, Y. -K. Lee, W. -J. Chun, J. Catal., 221 (2004) 263.

26. H. Y. Zhao,D. Li,P. Bui,S. T. Oyama, Appl. Catal. A: Gen., 391 (2011) 305.

27. V. M. L. Whiffen, K. J. Smith, Energ. Fuel, 24 (2010) 4728.

28. X. Duan,Y. Teng,A. Wang,V. M. Kogan,X. Li,Y. Wang, J. Catal., 261 (2009) 232.

29. T. I. Korányi, Z. Vít, D. G. Poduval, R. Ryoo, H. -S. Kim, E. J. M. Hensen, J. Catal., 253

(2008) 119.

30. H. Loboue,C. Guillot-Deudon,A. Florin Popa,A. Lafond,B. Rebours,C. Pichon,T.

Cseri,G. Berhault,C. Geantet, Catal. Today, 130 (2008) 63.

31. Q. Guan, W. Li, M. Zhang, K. Tao, J. Catal., 263 (2009) 1.

32. G. Shi ,J. Shen, J. Mater. Chem. , 19 (2009) 2295.

107

33. C. L. Yaws,P. K. Narasimhan,C. Gabbula. Yaws’ handbook of antoine coefficients for

vapor pressure. 2009 [cited 2012 10/04]; Available from: http://www.knovel.com.

34. C. Stinner,Z. Tang,M. Haouas,Th. Weber,R. Prins, J. Catal., 208 (2002) 456.

35. A. Wang,L. Ruan,Y. Teng,X. Li,M. Lu,J. Ren,Y. Wang,Y. Hu, J. Catal., 229 (2005) 314.

36. J. A. Rodriguez,J. Y. Kim,J. C. Hanson,S. J. Sawhill,M. E. Bussell, J. Phys. Chem. B, 107

(2003) 6276.

37. A. Motesinos-Castellanos,T.A. Zepeda,B. Pawelec,E. Lima,J.L.G. Fierro,A. Olivas,J.A.

de los Reyes, Appl. Catal. a: Gen., 334 (2008) 330.

38. K. A. Layman ,M. E. Bussell, J. Phys. Chem. B, 108 (2004) 10930.

39. Z. Feng,C. Liang,W. Wu,Z. Wu,R. A. van Santen,C. Li, J. Phys. Chem. B, 107 (2003)

13698.

40. A. W. Burns,K. A. Layman,D. H. Bale,M. E. Bussel, Appl. Catal. A: Gen., 343 (2008)

68.

41. S. J. Sawhill, D. C. Phillips, M. E. Bussell, J. Catal., 215 (2003) 208.

42. S. J. Sawhill, K. A. Layman, D. R. Van Wyk, M. H. Engelhard, C. Wang, M. E. Bussell, J.

Catal., 231 (2005) 300.

43. Briggs, D., Practical surface analysis by auger and x-ray photoelectron spectroscopy, ed.

M.P. Seah. 1983, New York: Wiley.

44. J. F. Moulder,W. F. Stickle,P. E. Sobol,K. D. Bomben,J. Chastain, Handbook of x-ray

photoelectron spectroscopy. 1992, Eden Prairie, MN: Perkin Elmer.

45. G. E. McGuire,G. K. Schweitzer,T. A. Carlson, Inorg. Chem., 12 (1973) 2450.

46. F. Notheisz ,M. Bartok, J. Catal., 71 (1981) 331.

47. U. Gennari, R. Krammer, H. L. Gruber, Appl. Catal., 11 (1984) 341.

48. U. Gennari, R. Krammer, H. L. Gruber, Appl. Catal., 44 (1988) 239.

49. P. B. Weisz ,C. D. Prater, Adv. Catal. Relat. Subj., 6 (1954) 143.

50. D. Li,P. Bui,H. Y. Zhao,S. T. Oyama,T. Dou,Z. H. Shen, J. Catal., 290 (2012) 1.

108

Chapter 4

Kinetic and Spectroscopic Study of Deoxygenation of 2-

Methyltetrahydrofuran over Supported Nickel Phosphide

Catalyst

4.1. Introduction

The general conclusion from reaction network studies of oxygenated aromatic

compounds is that the aromatic ring must be hydrogenated at least partially prior to ring opening

(Chapter 1), indicating saturated cyclic ethers as a potential intermediate in hydrodeoxygenation.

Previous chapters presented the successful preparation of silica supported transition metal

phosphides among which Ni2P/SiO2 showed high reactivity towards hydrodeoxygenation of

ethanol and 2-methyltetrahydrofuran. In this chapter, attention is placed on investigating the

effect of the reactants (2-MTHF and H2) partial pressure on the HDO reactions and monitoring

the surface species involved in the reactions.

4.2. Materials and Experiments

4.2.1. Materials and synthesis

The supported nickel phosphide catalyst was synthesized via a phosphate precursor as

reported earlier (Chapter 3). In short, nickel (II) nitrate (Ni(NO3)2.6H2O (Alfa Aesar, 99.6%))

109

was added into an ammonium hydrogen phosphate ((NH4)2HPO4 (Aldrich 99%)) solution with a

2:1 metal to phosphorous ratio. The resulting solution was impregnated onto a silica support

(Cab-osil ® EH5 supplied by Cabot Corp) using the incipient wetness method. The mixture was

dried at 120 oC for 4 h and calcined at 600

oC for 6 h to obtain the final phosphate precursor.

Then, a temperature programmed reduction in a H2 stream (1 dm3 min

-1 of H2 per gram of

precursor)was carried out to obtain the corresponding phosphide. The temperature was increased

from 25 oC to 580

oC at a rate of 2

oC min

-1. Finally, the catalyst was cooled to 25

oC in a He

stream and was passivated in a 0.5 % O2/He stream for 2 h.

4.2.2. General characterization

General characterization techniques such as X-ray diffraction, carbon monoxide (CO)

chemisorption and Brunauer Emmett Teller (BET) surface area measurements were carried out to

confirm the presence of nickel phosphide and to quantify the active sites on the catalyst.

4.2.3. Low temperature reactivity test

The reactivity test was carried out in a cylindrical quartz reactor packed with 2.9 g of

catalyst – equivalent to 338 µmol of active site on the basis of CO uptake. Prior to the test, the

supported nickel phosphide was reactivated in H2 as temperature increased at a rate of 5 oC min

-1

from 25 oC to 450

oC and held at this temperature for 2 h. The reactivity experiment was carried

out at atmospheric pressure and temperatures ranging from 150 to 225 oC. The reactant was fed

to the catalyst bed in a H2 stream (100 cm3 (NTP) min

-1) flowing first through a saturator

110

containing a solution of 95 volume % of 2-MTHF and 5 volume % of heptane as an internal

standard. The saturator was kept at 0 oC to ensure a stable feed concentration. The feed

composition was obtained as follows by applying the Antoine equation to obtain the vapor

pressures of 2-MTHF and heptane and then using Raoult’s law for liquid mixtures. The Antoine

equation has the form logP4L' = A − BT+C and the coefficients A, B, and C are 7.13891, 1339.48,

and 234.353 respectively for 2-MTHF and 7.04605, 1341.89, and 223.733 for heptane [1] and

gave a saturated vapor pressure at 0 oC (saturator temperature) of P

sat 26.5 mmHg for 2-MTHF

and 11.2 mmHg for heptanes. The total vapor pressure of the mixture was 26 mmHg. At

atmospheric pressure, the gas-phase concentration of the reactant 2-MTHF in H2 is 3.3 molar %.

The space velocity and contact time were 667 h-1

and 149 s respectively and were calculated by

the following equations:

GHSV = Feed volume flow rate

Volume of the catalyst bed in which feed volume flow rate is 6000 [cm

3 (NTP) h

-1] and

volume of the catalyst bed is 9 cm3.

Contact time = Quantity of active sites of catalyst loaded

Reactant molar flow rate

The products were monitored by an online gas chromatograph (SRI 8610B) equipped

with an HP-1 100m x 0.25mm capillary column and a flame ionization detector. The reactants

and products were identified by comparing their retention times with those of commercial

standards and confirmed by gas chromatography – mass spectrometry (GC-MS) (Hewlett –

Packard, 5890-5927A).

111

4.2.4. Kinetic test

The kinetic test was carried out in a trickle bed reactor at 300 oC and a total pressure of 1

atm. The reactant solution (a liquid mixture of 2-MTHF and 5 vol % of heptane) was delivered

into the reactor via a vaporizer kept at 300 oC. A mixed flow of H2 and He (total flow rate of 127

cm3

(NTP) min-1

) brought the vaporized reactant to the reactor chamber. The catalyst bed in the

middle of the reactor chamber was made with 100 mg of supported nickel phosphide well mixed

in 1.5 g of quartz with the same pelletized size to a total volume of 2 mL. The kinetic study

consisted of two subsets of experiments. In the first one the partial pressure of 2-MTHF was kept

constant at 14.3 atm while the partial pressure of H2 was varied in the following order (in atm):

85 � 83 � 66 � 29 � 14 � 21 �47 �74. In the second experiment the partial pressure of H2

was kept constant at 67 atm while the partial pressure 2-MTHF was varied as follows (in atm):

4.4 � 18 � 31 � 34 �25 � 15 � 11 � 7.7. As in the reactivity test the products were

monitored via an online gas chromatograph (SRI 8610B) equipped with an HP-1 100m x

0.25mm capillary column and a flame ionization detector.

4.2.5. Temperature-programmed desorption

Temperature programmed desorption (TPD) of 2-MTHF in He (and H2) was conducted in

a tubular quartz reactor using 0.50 g of catalyst. The catalyst underwent the same reactivation as

discussed above and was cooled in He flow (100 cm3 (NTP) min

-1). At 30

oC and after a 20 min

period of stabilization in He, the reactor was evacuated to a pressure of less than 0.1 mmHg, and

then was filled with saturated vapor of 2-MTHF to a pressure of 20 mmHg. Excess and lightly

adsorbed 2-MTHF was removed by a second evacuation. Then, the catalyst was flushed in a He

112

flow (100 sccm) for 30 min at 1 atm. Temperature programmed desorption in He (or H2) (50 cm3

(NTP) min-1

) started as temperature increased from 30 oC to 600

oC at a rate of 3

oC min

-1. The

desorption was monitored by a quadruple mass spectrometer. The monitored masses and their

relative intensity are summarized below.

Table 4.1: Masses monitored during temperature programmed desorption of 2-MTHF and

their relative intensity

Chemical m/z Relative

intensity

2-MTHF 71.0 100

41.0 51.8

42.0 47.9

2-Pentene 55.0 100

42.0 41.8

70.0 38.6

2-Pentanone 43.0 100.0

86.0 20.2

41.0 11.9

Pentanal 44.0 100

29.0 52.3

41.0 41.0

n-Pentane 43.0 100.0

42.0 78.3

41.0 50.5

1-Pentanol 42.0 100.0

55.0 65.1

70.0 51.0

2-Pentnaol 45.0 100.0

55.0 18.7

43.0 13.7

4.2.6. Fourier transform infrared (FTIR) spectroscopic experiments

The spectroscopic study was carried out using a Digilab Excalibur Series FTS 3000

spectrometer. The detachable reaction cell was positioned parallel to the beam line so that the

laser beam passed through the catalyst wafer prior to reaching the liquid N2 cooled mercury-

113

cadmium-telluride detector. The chamber had an outlet and inlet for gas flows and a

thermocouple connected to a temperature controller to monitor and control the sample

temperature. Potassium bromide windows were placed at the ends of the cell and perpendicular

to the beam line. The catalyst sample was made of 32.5 mg of finely ground Ni2P/SiO2 pressed

into a self-supporting wafer (13 mm in diameter). The catalyst was pretreated in H2 (100 cm3

(NTP) min-1

) at 450 oC for 2 h prior to experiments. Background spectra were taken as the

catalyst cooled down in either H2 or He (100 cm3 (NTP) min

-1). All spectra are shown with

subtraction of the background contribution. In the temperature programmed desorption

experiments, the catalyst sample was dosed with 0.32% 2-MTHF/He (100 cm3 (NTP) min

-1)

until saturation at 100oC and 1 atm. Then, sample was purged with carrier gas for 300 s to

remove the excessive 2-MTHF. During temperature programmed desorption spectra were

recorded at each 25 oC interval as temperature increased from 100 – 300

oC.

In the steady-state

experiments, 2-MTHF was fed at a constant concentration (0.32 mol %) in either H2 or He and

temperature was varied in the range of 105 - 300 oC at 50

oC interval, spectra were taken after the

catalyst was stabilized in 2-MTHF stream for 1.5 h. The transient experiment has the same

conditions with the steady state experiment; however, spectra were taken starting at the moment

2-MTHF was introduced.

4.3. Results and Discussion

4.3.1. Characterization

114

Fig. 4.1 shows the X-ray diffraction pattern for Ni2P/SiO2 exhibiting a broad peak at 20o

assigned to amorphous silica support and three sharp peaks at 40.8o, 44.6

o, and 47.3

o that

correspond well to the PDF pattern for Ni2P. The broad peak at 54.2o consists of an overlap of the

two peaks at 54.2o and 54.9

o suggested by the PDF pattern. XRD result confirms the presence of

Ni2P crystals in the catalyst. Furthermore, the crystal size of Ni2P particles can be deduced from

the width at half-height of the three most prominent peaks at 40.8o, 44.6

o, and 47.3

o (Table 4.2,

Supplementary Information).

20 30 40 50 60 70

Ni2P PDF 00-003-0953

Ni2P/SiO

2

2θ

Inte

nsity

/ a.u

.

Fig. 4.1 X-ray diffraction pattern of Ni2P/SiO2

Table 4.2 lists the results of general characterization done on the catalyst. These results

are in close agreement with the ones reported previously. Crystallite size was calculated using

the Scherrer equation Dc =pq

@3%4 r! in which K is the shape factor with a value of 0.9, λ is the

X-ray wavelength, β is the line broadening at half- maximum intensity in radians (accounting for

instrumental broadening of 0.1o), and θ is the Bragg angle. The equation was applied to the three

peaks at 40.8o, 44.6

o, and 47.3

o (see Supplementary Information) and the average value for

115

crystallite size is reported in Table 4.2. Measurements from CO chemisorption provided

estimation of the dispersion of metal sites following the equation:

Dispersion %! = CO uptake jμmol

gk

Loading l1.16µmolg

o100%

Table 4.2: Characterization results of Ni2P/SiO2; * BET surface area value of SiO2

Catalyst BET Surface Area

m2 g

-1 Crystallite Size

nm

CO uptake

µmol g-1

Dispersion

%

Ni2P/SiO2 177 (334*) 14.5 115 10

4.3.2. Low temperature reactivity test

The reactivity of 2-MTHF on Ni2P/SiO2 at low temperatures is presented as an Arrhenius

plot of 2-MTHF conversion over Ni2P/SiO2 (Fig. 4.2). A large catalyst loading (2.9 g) and a long

contact time (149s) between the reactant and the active sites were used to obtain conversion at

low temperatures and help improve the accuracy of the observed reaction activation energy. The

temperatures were varied in a downwards and upwards order as follows: 275-250-200-175-150-

162-187-223-225 oC to ensure the data were not affected by deactivation. The result exhibits an

excellent linear relationship between the logarithm of the turnover frequency and the inverse

temperature. The apparent activation energy for 2-MTHF conversion on Ni2P/SiO2 from the

slope of the plot was 134 kJ/mol. This value is slightly higher than the one reported previously

because the reaction rate starts to deviate from Arrhenius temperature dependency as temperature

116

increases past 225 oC yet a previous study started from 250

oC [Chapter 3]. Furthermore, in the

previous study there were less data points available to establish an accurate value. Kreuzer and

Kramer reported an apparent activation energy value of 75kJ/mol for the conversion of

tetrahydrofuran on silica supported Pt [2]. The higher value from 2-MTHF conversion indicates

a more complex reaction network for this molecule.

2.0 2.1 2.2 2.3 2.4

1E-6

1E-5

1E-4

Ni2P/SiO

2

Eactivation

=137.8 kJ/mol

225oC

9.5%

223oC

8.33%200

oC

1.78%

187oC

0.65%

175oC

0.25%

162oC

0.09%

150oC

0.03%

2-M

TH

F T

urn

ove

r F

req

ue

ncy /

s-1

Inverse Temperature / 10-3 K

-1

Fig. 4.2. Arrhenius plot of 2-MTHF conversion over Ni2P/SiO2

4.3.3. Kinetic study

Previous result on 2-MTHF reactions on Ni2P/SiO2 showed that pentenes and 2-

methylfuran are primary products, 2-pentanone and n-pentanal are secondary products, and

pentane and butane are final products [Chapter 3]. The dependence of 2-MTHF conversion on

the partial pressures of H2 and the reactant itself was investigated via a kinetic study (Table 4.3).

In the first series 2-MTHF partial pressure was held constant while H2 pressure was varied and in

117

the second series the opposite was carried out. In both experiments, total conversion and turnover

frequency increase as the concentration of either H2 or 2-MTHF increases. However, the

conversion of 2-MTHF shows a larger dependence (~ 3 times) on the partial pressure of 2-MTHF

than H2. As 2-MTHF pressure decreases 77 % from 34.1 to 7.7 kPa, the rate decreases 81 % from

0.75 to 0.14 s-1

; meanwhile, as H2 pressure decreases 83 % from 83.9 to 14.5 kPa, the rate of 2-

MTHF converted decreases 28 % from 0.32 to 0.23 s-1

.

Table 4.3 Effects of H2 and 2-MTHF partial pressures on the conversion of 2-MTHF over

Ni2P/SiO2

Partial Pressure

kPa

Total

Conversion

%

TOF

mol

mol-1

active

site s-1

Product Selectivity

%

H2 2-MTHF C4 Pentenes Pentane 2-

methylfuran

2-

pentanone

83.9 14.5 56.9 0.32 22.4 0.9 51.4 5.4 18.9

75.1 14.5 55.2 0.31 22.2 1.0 42.2 7.3 25.6

67.0 14.5 54.3 0.30 20.8 0.9 42.5 6.7 27.5

47.8 14.6 50.3 0.28 18.7 1.2 35.6 9.9 33.6

29.0 14.5 49.0 0.27 15.4 1.6 24.0 16.9 41.0

20.9 14.5 43.8 0.24 16.8 2.8 21.8 23.1 34.8

14.5 14.6 41.8 0.23 16.2 6.6 13.3 36.6 26.7

67.0 34.1 55.4 0.75 20.3 1.5 32.0 13.9 30.6

67.0 30.9 54.7 0.65 19.6 1.5 32.6 13.2 31.0

67.0 24.8 55.3 0.53 20.4 1.2 33.3 11.9 31.7

67.0 17.9 55.5 0.38 18.7 1.3 44.3 10.0 24.8

67.0 14.5 54.3 0.30 20.8 0.9 42.5 6.7 27.5

67.0 10.7 50.8 0.21 22.2 1.0 44.1 7.9 24.2

67.0 7.7 48.5 0.14 20.6 1.1 37.9 10.2 29.5

The selectivity of products is presented in Table 4.3 and Fig. 4.3. In the experiment where

H2 partial pressure was varied, a trend similar to contact time study was observed. As H2

concentration increases, the selectivity towards unsaturated products such as 2-methylfuran and

118

pentenes decreases, the selectivity towards the secondary product of 2-pentanone reaches a

maximum, the selectivity towards saturated final product of pentane increases significantly, and

the selectivity towards the cracking product of butane rises slightly. Note that the reaction is not

mass transfer limited as the Weisz-Prater coefficient is less than ; thus, the concentrations in the

immediate vicinity of the active sites are the same in the bulk. Considering the experimental

conditions which involve an excessive amount of catalytic active sites (11.5 µmol CO uptake)

available for the reaction of a constant reactant flow (6.4 µmol s-1

), the observed selectivity

profiles hint that the surface reaction may be the rate limiting step in the reaction mechanism.

In the experiment where 2-MTHF partial pressure was varied, total conversion and

selectivity of the products do not show a significant change as in the H2 experiment (Fig. 4.3).

The results show two steady regions below and above 20 kPa of 2-MTHF. As 2-MTHF flow

increases past 7.9 µmol s-1

(corresponding to 18.5 kPa), the selectivity towards the final product

– pentane shifts to a lower steady value, the selectivity towards intermediate products – 2-

pentanone, 2-methylfuran, and pentenes shifts to higher values while butane stays relatively the

same. The cracking product maintains a steady selectivity in both experiments as well as in the

contact time study (Fig. 4.3). If either the adsorption of 2-MTHF or the desorption of the

products was rate-limiting then the surface reaction would be fast, resulting in a high and steady

selectivity towards pentane and low selectivity towards other intermediate products. The case

was not observed here. Therefore, the surface reaction is likely the rate limiting step for the HDO

mechanism.

119

5 10 15 20 25 30 350

10

20

30

40

50

60

Se

lectivity /

%

P2MTHF

/ kPa

10 20 30 40 50 60 70 80 900

10

20

30

40

50

60

Total Conversion

Pentane

2-Pentanone

Butane

2-Methylfuran

Pentenes

PH

2

/ kPa

Fig. 4.3. Selectivity of products in partial pressure experiment

The apparent order of 2-MTHF conversion with respect to H2 and 2-MTHF was obtained

using the simple power rate law where TOF is the turnover frequency with the unit of mole of 2-

MTHF converted per mole of active sites per second, k is the overall rate constant, �<���� and

��Bare the partial pressures of 2-MTHF and H2, and a and b are the orders of the reaction with

respect to 2-MTHF and H2, respectively.

��� = ��<����� ��B�

The power rate law could be presented in a linear form as:

�� ���! = ln �! + ��� �<����! + ��� ��B!

120

6 7 8 9 10 20 30 400.1

T

OF

/ m

ol m

ol-1

active

site

s s

-1

P2MTHF

/ kPa

20 40 60 80 1000.2

0.3

PH

2

/ kPa

Fig. 4.4. Estimate of reaction order of H2 and 2-MTHF

Overall reaction orders a and b are determined from the slopes in Fig. 4.4. The

conversion of 2-MTHF on Ni2P/SiO2 has the order of 1.07 with respect to 2-MTHF and 0.17

with respect to H2. The first order of 2-MTHF is as expected. The low reaction order of H2 may

be due to the dehydrogenation reaction to form 2-methylfuran. At low H2 partial pressures (< 30

kPa), the conversion of 2-MTHF produces more 2-methylfuran, pentenes, and 2-pentanone than

the saturated pentane. It is likely that the adsorbed species donate hydrogen to active sites and

later on utilize the hydride species for their transformations, thus reducing the dependence on the

gas phase hydrogen. This is consistent with the excellent hydride transfer capacity of nickel

phosphides. On the other hand, extra hydrogen is necessary to form the final product – pentane.

121

4.3.4. Temperature programmed desorption

100 200 300 400

in He flowa)M

ass I

nte

nsity

/ a

.u.

Temperature / oC

100 200 300 400

in H2 flow b)

Masses

43

71

41

42

29

45

86

55

44

Fig. 4.5: Temperature programmed desorption of 2-MTHF on Ni2P/SiO2 in H2 and He at

atmospheric pressure, heating rate of 3 oC min

-1, injection temperature was at 30

oC

Spectra of masses 29, 41, 42, 43, 44, 45, 55, 71, and 86 in H2 flow (a) and He flow (b)

during temperature programmed desorption were shown (Fig. 4.5). Main peaks centered at 87 oC

were detected in both He and H2 flow. These peaks indicate desorption of the physisorbed 2-

MTHF. At around 210 oC a small peak was detected for masses 43, 42, 41 and 29 in H2 flow but

not in He flow. These masses could originate from either pentane, 2-pentanone or n-pentanal.

Additional TPD was carried out with a focus on masses 43, 42, 41, and 71. An elevated injection

temperature of 150 oC was applied to eliminate the physisorption of 2-MTHF. The results are

shown in Fig. 4.6. Here, mass 71 has no peak, indicating that this is not 2-MTHF. Masses 43, 42,

122

and 41 exhibit clear peaks at 210 oC and are consistent with the previous TPD experiment (Fig.

4.5). Relative intensity of each mass was calculated by taking the area of the desired mass

divided by the area of mass 43 and multiplied by 100. The calculation gives 100, 60, and 60 as

relative intensity of mass 43, 42, and 41 respectively. These values are closest to the value listed

for n-pentane in Table 4.1. The deviation from the reference values may be due to the very small

amount of pentane produced.

200 300 400 500

Mass Inte

nsity

/ a.u

.

Temperature / oC

Mass

43

41

42

71

Fig. 4.6 Temperature programmed desorption of 2-MTHF on Ni2P/SiO2 in H2 at

atmospheric pressure, heating rate of 3 oC min

-1, injection temperature was at 150

oC

4.3.5. Fourier transform spectroscopy

In order to observe the catalyst surface during reaction, a series of infrared spectroscopy

measurements were carried out at conditions of temperature programmed desorption, steady-

state reaction, and transient reaction. All spectra shown here include background subtraction at

the temperature of the measurement. Table 4.4 lists the main peaks and their assignments.

123

Temperature programmed desorption of 2-MTHF on the pure SiO2 support is shown in Fig. 4.7.

The high wavenumber region in the left panel shows a peak at 3746 cm-1

, a broad feature from

3500 to 3100 cm-1

, and a cluster of peaks at 2978, 2939, and 2880 cm-1

. The first two are

contributions from hydroxyl group associated with the silica support [3-5] and hydrogen-bond

interactions, and the last is from C-H stretching of the methyl and methylene groups of 2-MTHF

[6]. There seems to be interactions between 2-MTHF and the SiOH group due to the negative

feature surrounding the peak at 3746 cm-1

. The lower wavenumber region in the right panel

shows peaks at 1464 cm-1

with a shoulder at 1450 cm-1

, and a doublet at 1385 and 1360 cm-1

,

attributed to C-H bending vibrations (δs CH2, δas CH3 and δ CH for secondary C) of 2-MTHF [7].

Slight features at 1070 and 1000 cm-1

are from either C-O-C or C-C-C stretching of 2-MTHF.

The results showed that 2-MTHF is adsorbed weakly on pure the SiO2 support, probably via

hydrogen bonding with the silanol group. As temperature increases 2-MTHF simply desorbs

from the inert support without any chemical conversion; no features were observed for any

products. For this reason, the same experiment in He was not carried out.

124

4000 3500 3000

2880

2939

2978

3746

Si-OH

0.25

105 oC

125 oC

150 oC

175 oC

200 oC

250 oC

300 oC

Ab

so

rba

nce

Wavenumber / cm-1

2500 2000 1500 1000

1357

1450

1464

1385

0.25

Fig. 4.7. Temperature programmed desorption of 2-MTHF from the support SiO2 in H2

Considerable changes occur in the FTIR spectra as nickel phosphide is incorporated to

the support (Fig. 4.9). As 2-MTHF adsorbs on and desorbs from the catalyst, the following peaks

were noticed:

1/ Two negative peaks at 3745 and 3667 cm-1

are attributed to the consumption of the

hydroxyl groups associated with the support and phosphorus. Their negative intensity decreases

as temperature increases; eventually the peaks turn positive indicating a recovery of the

corresponding hydroxyl species. The silanol groups on the SiO2 are not acidic enough even for

the protonation of 2-methylpiperidine – a basic compound [8],therefore it is likely that silanol

group does not protonate 2-MTHF here. Instead, the P-OH group is likely to be involved here. A

previous FTIR analysis of P2O5/SiO2 showed the P-OH vibration at 3660 cm-1

. This band is close

to that of Ni2P/SiO2 catalyst. It was proposed that the P-OH groups at 3668 cm-1

are associated

125

with the Ni2P particles (Fig. 4.8) and the oxygen in these hydroxyl groups probably came from

incompletely reduced phosphate species [8].

Fig. 4.8 Proposed structural models of (a) P2O5/SiO2 and (b) Ni2P/SiO2 [8]

2/ An overlapped signal at 2978, 2939, and 2885 cm-1

is assigned to aliphatic C-H

stretching that comes from the adsorbed reactant and products.

3/ A peak at 2450 cm-1

originates from P-H stretching of phosphite species [7]. Unlike the

trend observed in the PO-H stretching (3667 cm-1

) which continues rising even as all species

desorbed, the P-H peak (2450 cm-1

) diminishes along with the desorption of surface species,

suggesting some hydrogen transfer activity between the adsorbed species and the phosphite

species.

4/ A peak at 2066 cm-1

that starts at 200 oC and reaches a maximum at 250

oC. This peak

belongs to adsorbed CO species, a decarbonylation product. The peak position matches with the

Ni0- CO complex [9, 10], indicating that the decarbonylation reaction takes place near or on Ni

metal sites – a result observed in previous studies [11, 12].

126

5/ A broad region contains peaks at 1693 and 1634 cm-1

in which the former is assigned

to a C=O stretch from adsorbed ketone species and the latter is assigned to a C=C stretch from

adsorbed alkenes [7]. The accompanying alkene C-H stretching vibration, usually appears at

3015-3100 cm-1

does not appear here probably because of overlap with the alkane C-H stretch

region which starts at 3030 cm-1

(Fig. 4.9). Another possible explanation is that the alkene

species has a π-interaction on the active sites; this π-absorption may stabilize the species and

lower the frequency of the alkene C-H stretch into the alkane C-H stretch region, resulting in

being masked by the predominant alkane C-H vibrations. The intensity of both C=O and C=C

stretches quickly decline with temperature.

6/ Two peaks at 1464 and 1385 cm-1

are contributed from the alkane C-H bend of 2-

MTHF as these were also observed over pure SiO2. The peak at 1464 cm-1

has a small left

shoulder at 1488 cm-1

and a right shoulder at 1450 cm-1

. The right shoulder is due to the CH2

scissoring vibration of the non-aromatic ring of 2-MTHF. Cyclization is known to decrease the

frequency of CH2 bending vibrations while having little effect on C-H stretching vibrations for

unstrained cyclic hydrocarbons [7]. For example, cyclohexane absorbs at 1452 cm-1

while n-

hexane absorbs at 1468 cm-1

[7]. In the 1385-1360 cm-1

region, there is one peak at 1385 cm-1

,

attributed to C-H bending of the methyl group of 2-MTHF. On pure silica (Fig. 4.7), the same

region showed a doublet (1385 and 1357 cm-1

) which is a result of the interaction between the in

phase and out-of-phase CH3 bending of the two methyl and methylene groups attached to the

secondary carbon atom of 2-MTHF. Therefore, if 2-MTHF was absorbed in a η1 (O) mode on an

active site, there should present a doublet as seen on pure silica (Fig. 4.7). So, the loss of the

peak at 1360 cm-1

on Ni2P/SiO2 suggests that 2-MTHF may have adsorbed onto the catalyst in a

parallel η5 conformation in which the methyl or methylene group attached to the secondary

127

carbon atom was immobilized. As this phenomenon was only observed on Ni2P/SiO2, the

bonding must have occurred on the Ni2P active phase or at the catalyst/support boundary.

4000 3000

0.1

P-OH

1634

2450 103

oC

125oC

150oC

175oC

201oC

226oC

251oC

275oC

300oC

325oC

2939

2066

3667

14641385

1693

29782885

3745

Si-OH

A

bso

rba

nce

Wavenumber / cm-1

2000 1000

0.1

Fig. 4.9. Temperature programmed desorption of 2-MTHF from Ni2P/SiO2 in H2

Overall, the result indicates some transformation of adsorbed 2-MTHF at a temperature

as low as 103 oC. The presence of 2-pentanone species (#5) in the low temperature range

compared to the CO species in the high temperature range indicates that ring-opening has a

lower activation barrier energy than decarbonylation.

A significant change was observed as the temperature-programmed desorption was

carried out in He (Fig. 4.10). In the high wavenumber region the FTIR spectra show a less

significant recovery of the SiOH and POH groups at 3746 and 3667 cm-1

and an elimination of

the P-H species at 2450 cm-1

(Fig. 4.10). In the low wavenumber region the features at 1693 and

1636 cm-1

are present, yet follow a different trend. The νC=C band (1636 cm-1

) is more intense

initially and the νC=O band (1693 cm-1

) is less prominent. As temperature increases the νC=C band

128

decreases and broadens to lower wavenumbers while the νC=O band rises to a maximum and then

decreases. As temperature passes 275 oC two new bands appeared at 1656 and 1600 cm

-1,

corresponding to conjugated C=C stretching [7]. These features maybe attributed to pentadienes

or furan derivatives. A small but clear peak at 1488 cm-1

is seen from 105oC to 200

oC. A previous

study on thiophene assigned the band at 1491 cm-1

to a di-π-absorbed 1,3-butadiene species and

the band at 1449 cm-1

to both π and σ-bonded butenes [13]. It might be the case here with

pentadiene as a product. Other peaks associated with pentadiene in the 3000 cm-1

region were not

observed, due to the small quantity and overlap with the alkane C-H region as explained above

(#6).

4000 3000

0.1

1636

1600

105oC

125oC

150oC

175oC

200oC

225oC

250oC

275oC

300oC

325oC

14641385

2939

3667

2978

2885

3746

Ab

so

rba

nce

Wavenumber / cm-1

2000 1000

0.1

1693

Fig. 4.10 Temperature programmed desorption of 2-MTHF from Ni2P/SiO2 in He

Overall, the initial trend of νC=C and νC=O species is in agreement with the contact time

study in which pentenes decreases over time and 2-pentanone reaches a maximum before

129

declining. The presence of conjugated νC=C at the high temperature indicates the conversion of 2-

MTHF in He proceeds via dehydrogenation route.

The reaction of 2-MTHF on Ni2P/SiO2 was also monitored in H2 and He using FTIR (Fig.

4.11, Fig. 4.12). In the presence of constant 2-MTHF flow, additional vibration absorptions at

1100 and 1025 cm-1

was observed. These peaks correspond to the slight feature at 1067 and 1000

cm-1

observed in desorption experiments and are attributed to the C-O-C or C-C-C stretching of

2-MTHF [7]. The shift to higher frequency occurs because of the increased coverage of

adsorbed species. The increase in temperature which lowers the intensity of other bands seems

to not affect these bands, especially the band at 1100 cm-1

. The spectra of 2-MTHF conversion in

H2 are similar to those observed in the desorption experiment under H2 (Fig. 4.11, Fig. 4.9). The

main difference shows at the 1700-1600 cm-1

region where the νC=O stays at the same location

while the νC=C starts at lower wavenumber – 1605 cm-1

. The broad C=C band at 1605 cm-1

indicates the formation of a conjugated structure which could be 2-methylfuran, pentadiene, or a

mixture of 2-methyltetrahydrofuran and pentenes species. As temperature increases, both C=O

and C=C stretching absorption decrease their intensity. This is as expected since reaction

proceeds faster at higher temperature, hence the less chance to observe intermediate species on

the surface. During the decrease, the C=O band stays at the same location, whereas the C=C

band starts at 1605 cm-1

and moves toward higher frequency, indicating the diminishing of

conjugated alkenes and rising of mono alkenes. The ketone species appeared to decline at a

slower rate than the alkene species, consistent with the contact time study in which 2-pentanone

is the secondary product and pentene is the primary product. Carbon monoxide was detected at

200 oC, the same temperature as in the previous desorption experiment (Fig. 4.9). The intensity

130

of absorbed CO goes through a maximum at 250 oC, again, the same as the previous result Fig.

4.9.

4000 3000 2000 1000

125 oC

νC=O

1449

10251100

243513851464

16051693

0.25

aliphatic νC-H

150 oC

105 oC

250 oC

300 oC

200 oC

3746

Si-OH

3668

P-OH

2880

29392978

Ab

so

rba

nce

Wavenumber / cm-1

Fig. 4.11. FTIR spectra of 2-MTHF adsorbed on Ni2P/SiO2 in 0.32% 2-MTHF/H2 flow as a

function of temperature

In 0.32% of 2-MTHF in He, similar species to the H2 experiment was detected, yet the

trend of these species exhibits dramatic change (Fig. 4.12). First, a slower recovery of SiOH and

POH groups at 3746 amd 3667 cm-1

was observed as these peaks remain negative as temperature

increases. Second, there is no P-H at 2450 cm-1

. At low temperature, there are C=O and C=C

bands at 1690 and 1636 cm-1

, respectively. The intensity of these peaks decreases as temperature

increases. As temperature passes 150 oC a new set of peaks forms at 1656 cm

-1 and 1600 cm

-1.

The new peaks are attributable to conjugated alkene species such as 2-methylfuran and

pentadienes. The intensity of the peak at 1600 cm-1

increases significantly as temperature

131

increases, suggesting that the conversion of 2-MTHF in He proceeds mainly via

dehydrogenation.

4000 3000 2000 1000

0.25

105oC

125oC

150oC

200oC

250oC

300oC

1600

1025

1100

14491385

1693

2939

14643668

P-OH

2978

2880

3746

Si-OH

Ab

so

rba

nce

Wavenumber / cm-1

Fig. 4.12. FTIR spectra of 2-MTHF adsorbed on Ni2P/SiO2 in 2-MTHF/He flow as a

function of temperature

A transient study in H2 was conducted by monitoring the catalyst surface as 2-MTHF was

introduced into the system. Fig. 4.13 shows spectra as time increases from 0 to the steady-state

of 150 min. The results are in good agreement with the result of the contact time study. Alkenes

were formed first with initial growth of the C=C vibration at 1630 cm-1

. Then, 2-pentanone

followed with a C=O band at 1693 cm-1

starting at 90s and growing past the C=C band after 300

s. At steady state, the surface intermediates came to equilibrium with each other and the intensity

of the ketone species remained higher than the alkene species.

132

2000 1750 1500

112 s

96 s

80 s

64 s

48 s

32 s

16 s

0 s

1693

νC=O

1630

νC=C

13851449

1466

Ab

sorb

an

ce

Wavenumber / cm-1

2000 1750 1500

150 m

90 m

50 m

1630-1600

νC=C

1693

νC=O

5.3 m

3.2 m

2.9 m

2.7 m

2.4 m

2 m

Fig. 4.13. Transient FTIR spectra following the introduction of 2-MTHF onto Ni2P/SiO2 in

H2

Table 4.4: Assignment of FTIR peaks

Wavenumber / cm-1

Assignment Contributions Reference

3746 SiO-H stretch Si-OH [3-5]

3668 PO-H stretch P-OH [3-5]

2978, 2939, 2885 Aliphatic C-H

stretch

-CH3, -CH2, -CH, 2-MTHF,

2-pentanone, pentenes, 2-

methylfuran

[7]

2450 P-H stretch Phosphite residue [14]

2066 C-O CO [5, 15]

1693 C=O stretch 2-Pentanone, n-pentanal [7]

1634 C=C stretch,

unconjugated

Pentenes [7]

1653&1600 C=C stretch,

conjugated

Pentadienes, 2-methylfuran [7]

1490 C=C stretch Pentenes, pentadienes, 2-

methylfuran

[5]

1464&1450&1385 C-H bending -CH3, -CH2, -CH 2-MTHF,

pentenes, 2-pentanone,

pentenes, 2-methylfuran

[7]

1385 Tertiary C-H

(alkanes)

2-MTHF [7]

1100&1025 C-O stretch 2-MTHF, 2-methylfuran [7]

133

4.4. Conclusions

The hydrodeoxygenation of a heterocyclic O-compound (2-methyltetrahydrofuran) was

investigated on Ni2P/SiO2 using a combination of temperature programmed desorption, kinetic

study, and Fourier transform infrared spectroscopy. The apparent activation energy was

calculated to be 134 kJ/mol. The conversion of 2-MTHF was found to be first order in 2-MTHF

partial pressure and an order of 0.17 in H2 partial pressure. The low dependence on external

hydrogen indicates hydrogen transfer between the reactant and the catalyst active sites. FTIR

result on the Ni2P surface was in agreement with the contact time study in which an alkene

species was formed first followed by a ketone species, indicating pentenes are primary products

and 2-pentanone was a secondary product.

4.5. Supplementary Information

4.5.1. Crystallite size calculations

Crystallite size was calculated using the Scherrer equation Dc =pq

@3%4 r! in which:

K is the shape factor with a value of 0.9

λ is the X-ray wavelength with a value of 0.154178 nm

β is the line broadening at half the maximum intensity in radians (accounting for

instrumental broadening of 0.1o

θ is the Bragg angle

134

θ

Bragg angle

/ degree

Crystal

Face

β

Width at half height

/ rad

Crystallite size

/ nm

Average

crystallite size

/ nm

40.8 111 0.0103 14.4

14.5 44.6 201 0.0103 14.6

47.3 210 0.0105 14.5

4.5.2. Weisz – Prater criterion for internal diffusion

The Weisz – Prater criterion CWP uses measured values of the rate of reaction to determine if

internal diffusion is limiting the reaction as follows:

CWP = Actual reaction rate

A diffusion rate=−|} ~��!� ���<

.�g}�

Internal mass transfer effects can be neglected when the value of CWP is less than 1. The

parameters used for the criterion are listed in Table 4.5. The observed rates −|} ~��!� were

calculated at each contact time and listed along with the corresponding CWP values in Table 4.6.

Table 4.5. Parameters in the Weisz-Prater Criterion

R Catalyst particle radius / cm 0.08

ρc

(Ni2P)

Solid catalyst density for Ni2P /

g·cm-3

0.3

De Effective diffusivity/ cm2·s

-1 0.1

135

Table 4.6: CWP values for reactivity test at low temperature

Temperature

/ oC

CAs

Gas concentration at

the catalyst surface

/ mol·cm-3

−|} ~��!�

Reaction rate

/ mol g-1

s-1

CWP values

150 9.50E-07 2.42E-10 2.31E-16

162 9.24E-07 7.11E-10 6.57E-16

175 8.97E-07 1.92E-09 1.72E-15

187 8.74E-07 5.07E-09 4.43E-15

200 8.50E-07 1.38E-08 1.17E-14

223 8.11E-07 6.47E-08 5.25E-14

225 8.07E-07 7.38E-08 5.96E-14

References

1. C. L. Yaws,P. K. Narasimhan,C. Gabbula. Yaws’ handbook of antoine coefficients for

vapor pressure. 2009 [cited 2012 10/04]; Available from: http://www.knovel.com.

2. K. Kreuzer ,R. Kramer, J. Catal., 167 (1997) 391.

3. A. P. Legrand,H. Hommel,A. Tuel, Adv. Col. Interface Sci. , 33 (1990) 91.

4. S. T. Oyama, T. Gott, K. Asakura, S. Takakusagi, K. Miyazaki, Y. Koike, K. K. Bando, J.

Catal., 268 (2009) 209.

5. Y. -K. Lee ,S. T. Oyama, J. Catal., 239 (2006) 376.

6. M. Hawkins ,L. Andrews, J. Am. Chem. Soc. , 105 (1983) 2523.

7. R. M. Silverstein ,F. X. Webster, Spectrometric identification of organic compounds. 6th

ed. 1998, New York: Wiley.

8. S. T. Oyama ,Y. –K. Lee, J. Phys. Chem. B, 109 (2005) 2109.

9. Peri, J. B., Discuss Faraday Soc., 41 (1966) 121.

10. Peri, J. B., J. Catal., 86 (1984) 84.

11. R. H. Bowker, M. C. Smith, M. L. Pease, K. M. Slenkamp, L. Kovarik, M. E. Bussell,

ACS Catal., 1 (2011) 917.

12. S. Sitthisa ,D. Resasco, Catal. Lett., 141 (2011) 784.

13. Z. Wu,Z. Hao,P. Ying,Z. Wei,C. Li,Q. Xin, J. Phys. Chem. B, 104 (2000) 12275.

14. T. C. Stringfellow,J. D. Trudeau,T. C. Farrar, J. Phys. Chem., 97 (1993) 3985.

15. K. I. Hadjiivanov ,G. N. Vayssilov, Adv. Catal., 47 (2002) 307.

136

Chapter 5

Conclusions

5.1. Conclusions

The decomposition of ethanol in helium was investigated on Ni2P/SiO2 in comparison to

an acidic catalyst HZSM-5. Chemisorption measurements showed that Ni2P/SiO2 contained a

mixture of metal sites (134 µmol g-1

) and acidic sites (381 µmol g-1

) while HZM-5 had few metal

sites and a substantial amount of acidic sites (565 µmol g-1

). The nickel phosphide catalyst

exhibited a higher reactivity towards ethanol deoxygenation than the HZSM-5 catalyst.

To elucidate the decomposition mechanism of ethanol on Ni2P/SiO2, contact time

experiments were carried out. The results indicated that acetaldehyde was a primary product and

ethylene was a secondary product. A network of sequential reactions of surface intermediates

was proposed to explain the phenomenon. First, the reactant adsorbed on the catalyst surface,

then species reacted to form an adsorbed acetaldehyde species which could desorb or form

adsorbed alkene species; finally, the alkene species desorbed as ethylene. This rake-type

mechanism was supported by a simulation that produced good agreement with the experimental

data. Fourier transform infrared measurements confirmed the formation of the adsorbed aldehyde

and alkene intermediate species, thus, further validated the mechanism.

137

Reactions of 2-methyltetrahydrofuran on Ni2P/SiO2 were carried out on a series of

transition metal phosphides. The phosphides were prepared via the phosphate (A) precursors

(conventional method) and the phosphite (I) precursors (new method). The (A) catalysts went

through harsher preparation conditions and had lower surface area than the (I) catalysts. Both

type of catalysts showed comparable reactivity towards the HDO of 2-MTHF except for WP.

The difference was attributed to the surface P/W ratio that X-ray photoelectron spectroscopy

measurements showed to be twice as large for the (I) method than the (A) method.

At 300 oC and 1atm, reactivity towards 2-MTHF of the phosphides followed the order of:

Ni2P > WP > MoP > CoP > FeP > Pd/Al2O3. Iron group phosphides (Ni2P and CoP) produced

mostly pentane and butane, whereas group 6 metal phosphides (WP and MoP) and FeP produced

mostly pentenes and pentadienes.

For Ni2P/SiO2, contact time study showed pentene as primary products, 2-pentanone as

secondary product, and pentane as a final product. Fourier transform infrared spectroscopy

showed formation of surface intermediate species in consistency with the contact time study

results. The conversion of 2-MTHF on Ni2P showed a first order in 2-MTHF concentration and

an order of 0.17 in H2 concentration. The low hydrogen dependence indicated active hydrogen

transfer among the reactants, the products and the catalytic active sites.

5.2. Recommendations for future work

1/ Computational calculations

Based on the kinetic information from Chapter 4, one can propose various mechanisms of

2-MTHF reactions on Ni2P and perform numerical simulations to obtain the best fit mechanism

138

that describes the observed physical phenomenon. Density function theory is also an excellent

tool to predict adsorption modes of 2-MTHF that are most favored on Ni2P sites.

2/ Extended X-ray absorption fine structure spectroscopy (EXAFS)

In order to obtain more information on the active phase of Ni2P, extended X-ray

absorption fine structure spectroscopy can be used to characterize the catalyst before, during, and

after reaction. The technique reveals information on the bond distances (For examples Ni-Ni, Ni-

P, or Ni-O) present on the surface. Comparing the bonds formed during and after the reaction to

the ones before the reaction would give insight to the nature of active phosphide catalyst.

3/ Studies on WP catalysts

Steady state results showed that WP had little cracking products and favored the

formation of unsaturated carbons. Less cracking results in less CO formation and thus less

poisoning effect while unsaturated carbons may have great use in the synthetic polymer industry.

Therefore, it would be interesting to study the reactions on WP catalysts.